kvm_host.h 38.0 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
 * Bit 63 of the memslot generation number is an "update in-progress flag",
53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
 * 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.
 */
70
#define KVM_MEMSLOT_GEN_UPDATE_IN_PROGRESS	BIT_ULL(63)
71

72 73
/* Two fragments for cross MMIO pages. */
#define KVM_MAX_MMIO_FRAGMENTS	2
A
Avi Kivity 已提交
74

75 76 77 78
#ifndef KVM_ADDRESS_SPACE_NUM
#define KVM_ADDRESS_SPACE_NUM	1
#endif

79 80
/*
 * For the normal pfn, the highest 12 bits should be zero,
81 82
 * so we can mask bit 62 ~ bit 52  to indicate the error pfn,
 * mask bit 63 to indicate the noslot pfn.
83
 */
84 85 86
#define KVM_PFN_ERR_MASK	(0x7ffULL << 52)
#define KVM_PFN_ERR_NOSLOT_MASK	(0xfffULL << 52)
#define KVM_PFN_NOSLOT		(0x1ULL << 63)
87 88 89

#define KVM_PFN_ERR_FAULT	(KVM_PFN_ERR_MASK)
#define KVM_PFN_ERR_HWPOISON	(KVM_PFN_ERR_MASK + 1)
90
#define KVM_PFN_ERR_RO_FAULT	(KVM_PFN_ERR_MASK + 2)
91

92 93 94 95
/*
 * error pfns indicate that the gfn is in slot but faild to
 * translate it to pfn on host.
 */
D
Dan Williams 已提交
96
static inline bool is_error_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
97
{
98
	return !!(pfn & KVM_PFN_ERR_MASK);
X
Xiao Guangrong 已提交
99 100
}

101 102 103 104 105
/*
 * 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
106
static inline bool is_error_noslot_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
107
{
108
	return !!(pfn & KVM_PFN_ERR_NOSLOT_MASK);
X
Xiao Guangrong 已提交
109 110
}

111
/* noslot pfn indicates that the gfn is not in slot. */
D
Dan Williams 已提交
112
static inline bool is_noslot_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
113
{
114
	return pfn == KVM_PFN_NOSLOT;
X
Xiao Guangrong 已提交
115 116
}

117 118 119 120 121 122
/*
 * 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

123 124
#define KVM_HVA_ERR_BAD		(PAGE_OFFSET)
#define KVM_HVA_ERR_RO_BAD	(PAGE_OFFSET + PAGE_SIZE)
X
Xiao Guangrong 已提交
125 126 127

static inline bool kvm_is_error_hva(unsigned long addr)
{
128
	return addr >= PAGE_OFFSET;
X
Xiao Guangrong 已提交
129 130
}

131 132
#endif

133 134
#define KVM_ERR_PTR_BAD_PAGE	(ERR_PTR(-ENOENT))

135
static inline bool is_error_page(struct page *page)
136 137 138 139
{
	return IS_ERR(page);
}

140 141
#define KVM_REQUEST_MASK           GENMASK(7,0)
#define KVM_REQUEST_NO_WAKEUP      BIT(8)
142
#define KVM_REQUEST_WAIT           BIT(9)
143
/*
144 145
 * Architecture-independent vcpu->requests bit members
 * Bits 4-7 are reserved for more arch-independent bits.
146
 */
147 148 149 150
#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
151 152 153
#define KVM_REQUEST_ARCH_BASE     8

#define KVM_ARCH_REQ_FLAGS(nr, flags) ({ \
154
	BUILD_BUG_ON((unsigned)(nr) >= (FIELD_SIZEOF(struct kvm_vcpu, requests) * 8) - KVM_REQUEST_ARCH_BASE); \
155 156 157
	(unsigned)(((nr) + KVM_REQUEST_ARCH_BASE) | (flags)); \
})
#define KVM_ARCH_REQ(nr)           KVM_ARCH_REQ_FLAGS(nr, 0)
158

159 160
#define KVM_USERSPACE_IRQ_SOURCE_ID		0
#define KVM_IRQFD_RESAMPLE_IRQ_SOURCE_ID	1
161

162
extern struct kmem_cache *kvm_vcpu_cache;
A
Avi Kivity 已提交
163

164
extern spinlock_t kvm_lock;
165 166
extern struct list_head vm_list;

167 168 169 170 171 172
struct kvm_io_range {
	gpa_t addr;
	int len;
	struct kvm_io_device *dev;
};

173
#define NR_IOBUS_DEVS 1000
174

175
struct kvm_io_bus {
176 177
	int dev_count;
	int ioeventfd_count;
178
	struct kvm_io_range range[];
179 180
};

M
Marcelo Tosatti 已提交
181 182 183
enum kvm_bus {
	KVM_MMIO_BUS,
	KVM_PIO_BUS,
184
	KVM_VIRTIO_CCW_NOTIFY_BUS,
185
	KVM_FAST_MMIO_BUS,
M
Marcelo Tosatti 已提交
186 187 188
	KVM_NR_BUSES
};

189
int kvm_io_bus_write(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, gpa_t addr,
M
Marcelo Tosatti 已提交
190
		     int len, const void *val);
191 192 193 194
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);
195 196
int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
			    int len, struct kvm_io_device *dev);
197 198
void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
			       struct kvm_io_device *dev);
199 200
struct kvm_io_device *kvm_io_bus_get_dev(struct kvm *kvm, enum kvm_bus bus_idx,
					 gpa_t addr);
201

202 203 204 205 206 207 208 209 210 211
#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;
212
	bool   wakeup_all;
213 214 215 216
};

void kvm_clear_async_pf_completion_queue(struct kvm_vcpu *vcpu);
void kvm_check_async_pf_completion(struct kvm_vcpu *vcpu);
217
int kvm_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, unsigned long hva,
218
		       struct kvm_arch_async_pf *arch);
219
int kvm_async_pf_wakeup_all(struct kvm_vcpu *vcpu);
220 221
#endif

222 223 224
enum {
	OUTSIDE_GUEST_MODE,
	IN_GUEST_MODE,
225 226
	EXITING_GUEST_MODE,
	READING_SHADOW_PAGE_TABLES,
227 228
};

A
Avi Kivity 已提交
229 230 231 232 233 234 235 236 237 238
/*
 * 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;
};

239 240
struct kvm_vcpu {
	struct kvm *kvm;
241
#ifdef CONFIG_PREEMPT_NOTIFIERS
242
	struct preempt_notifier preempt_notifier;
243
#endif
244
	int cpu;
245
	int vcpu_id;
246 247
	int srcu_idx;
	int mode;
248
	u64 requests;
J
Jan Kiszka 已提交
249
	unsigned long guest_debug;
250

251 252 253
	int pre_pcpu;
	struct list_head blocked_vcpu_list;

254 255
	struct mutex mutex;
	struct kvm_run *run;
256

257
	int guest_xcr0_loaded;
258
	struct swait_queue_head wq;
259
	struct pid __rcu *pid;
260 261 262
	int sigset_active;
	sigset_t sigset;
	struct kvm_vcpu_stat stat;
W
Wanpeng Li 已提交
263
	unsigned int halt_poll_ns;
264
	bool valid_wakeup;
265

266
#ifdef CONFIG_HAS_IOMEM
267 268 269
	int mmio_needed;
	int mmio_read_completed;
	int mmio_is_write;
A
Avi Kivity 已提交
270 271 272
	int mmio_cur_fragment;
	int mmio_nr_fragments;
	struct kvm_mmio_fragment mmio_fragments[KVM_MAX_MMIO_FRAGMENTS];
273
#endif
A
Avi Kivity 已提交
274

275 276 277 278 279 280 281 282 283
#ifdef CONFIG_KVM_ASYNC_PF
	struct {
		u32 queued;
		struct list_head queue;
		struct list_head done;
		spinlock_t lock;
	} async_pf;
#endif

284 285 286 287 288 289 290 291 292 293 294 295
#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
296
	bool preempted;
297
	struct kvm_vcpu_arch arch;
298
	struct dentry *debugfs_dentry;
299 300
};

301 302
static inline int kvm_vcpu_exiting_guest_mode(struct kvm_vcpu *vcpu)
{
303 304 305 306 307 308
	/*
	 * 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();
309 310 311
	return cmpxchg(&vcpu->mode, IN_GUEST_MODE, EXITING_GUEST_MODE);
}

312 313 314 315 316 317
/*
 * 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
318 319 320 321
struct kvm_memory_slot {
	gfn_t base_gfn;
	unsigned long npages;
	unsigned long *dirty_bitmap;
322
	struct kvm_arch_memory_slot arch;
323
	unsigned long userspace_addr;
324
	u32 flags;
325
	short id;
A
Avi Kivity 已提交
326 327
};

328 329 330 331 332
static inline unsigned long kvm_dirty_bitmap_bytes(struct kvm_memory_slot *memslot)
{
	return ALIGN(memslot->npages, BITS_PER_LONG) / 8;
}

333 334 335 336 337 338 339
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);
}

340 341 342 343 344 345 346 347
struct kvm_s390_adapter_int {
	u64 ind_addr;
	u64 summary_addr;
	u64 ind_offset;
	u32 summary_offset;
	u32 adapter_id;
};

348 349 350 351 352
struct kvm_hv_sint {
	u32 vcpu;
	u32 sint;
};

353 354
struct kvm_kernel_irq_routing_entry {
	u32 gsi;
355
	u32 type;
356
	int (*set)(struct kvm_kernel_irq_routing_entry *e,
357 358
		   struct kvm *kvm, int irq_source_id, int level,
		   bool line_status);
359 360 361 362 363
	union {
		struct {
			unsigned irqchip;
			unsigned pin;
		} irqchip;
364 365 366 367 368 369 370
		struct {
			u32 address_lo;
			u32 address_hi;
			u32 data;
			u32 flags;
			u32 devid;
		} msi;
371
		struct kvm_s390_adapter_int adapter;
372
		struct kvm_hv_sint hv_sint;
373
	};
374 375 376
	struct hlist_node link;
};

377 378 379 380 381 382 383 384 385 386 387 388
#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

389 390 391 392
#ifndef KVM_PRIVATE_MEM_SLOTS
#define KVM_PRIVATE_MEM_SLOTS 0
#endif

393
#ifndef KVM_MEM_SLOTS_NUM
394
#define KVM_MEM_SLOTS_NUM (KVM_USER_MEM_SLOTS + KVM_PRIVATE_MEM_SLOTS)
395 396
#endif

397 398 399 400 401 402 403
#ifndef __KVM_VCPU_MULTIPLE_ADDRESS_SPACE
static inline int kvm_arch_vcpu_memslots_id(struct kvm_vcpu *vcpu)
{
	return 0;
}
#endif

404 405 406 407 408
/*
 * Note:
 * memslots are not sorted by id anymore, please use id_to_memslot()
 * to get the memslot by its id.
 */
409
struct kvm_memslots {
410
	u64 generation;
411
	struct kvm_memory_slot memslots[KVM_MEM_SLOTS_NUM];
412
	/* The mapping table from slot id to the index in memslots[]. */
413
	short id_to_index[KVM_MEM_SLOTS_NUM];
414
	atomic_t lru_slot;
415
	int used_slots;
416 417
};

A
Avi Kivity 已提交
418
struct kvm {
419
	spinlock_t mmu_lock;
420
	struct mutex slots_lock;
421
	struct mm_struct *mm; /* userspace tied to this vm */
422
	struct kvm_memslots __rcu *memslots[KVM_ADDRESS_SPACE_NUM];
R
Rusty Russell 已提交
423
	struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
424 425 426 427 428 429 430

	/*
	 * 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.
	 */
431
	atomic_t online_vcpus;
432
	int created_vcpus;
433
	int last_boosted_vcpu;
434
	struct list_head vm_list;
435
	struct mutex lock;
436
	struct kvm_io_bus __rcu *buses[KVM_NR_BUSES];
G
Gregory Haskins 已提交
437 438 439 440
#ifdef CONFIG_HAVE_KVM_EVENTFD
	struct {
		spinlock_t        lock;
		struct list_head  items;
441 442
		struct list_head  resampler_list;
		struct mutex      resampler_lock;
G
Gregory Haskins 已提交
443
	} irqfds;
G
Gregory Haskins 已提交
444
	struct list_head ioeventfds;
G
Gregory Haskins 已提交
445
#endif
446
	struct kvm_vm_stat stat;
447
	struct kvm_arch arch;
448
	refcount_t users_count;
449
#ifdef CONFIG_KVM_MMIO
450
	struct kvm_coalesced_mmio_ring *coalesced_mmio_ring;
451 452
	spinlock_t ring_lock;
	struct list_head coalesced_zones;
453
#endif
454

455
	struct mutex irq_lock;
456
#ifdef CONFIG_HAVE_KVM_IRQCHIP
457
	/*
458
	 * Update side is protected by irq_lock.
459
	 */
A
Arnd Bergmann 已提交
460
	struct kvm_irq_routing_table __rcu *irq_routing;
461 462
#endif
#ifdef CONFIG_HAVE_KVM_IRQFD
463
	struct hlist_head irq_ack_notifier_list;
464 465
#endif

466
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
467 468 469 470
	struct mmu_notifier mmu_notifier;
	unsigned long mmu_notifier_seq;
	long mmu_notifier_count;
#endif
471
	long tlbs_dirty;
472
	struct list_head devices;
473
	bool manual_dirty_log_protect;
474 475
	struct dentry *debugfs_dentry;
	struct kvm_stat_data **debugfs_stat_data;
476 477
	struct srcu_struct srcu;
	struct srcu_struct irq_srcu;
478
	pid_t userspace_pid;
A
Avi Kivity 已提交
479 480
};

481 482 483 484 485 486
#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__)
487 488 489
#define kvm_debug_ratelimited(fmt, ...) \
	pr_debug_ratelimited("kvm [%i]: " fmt, task_pid_nr(current), \
			     ## __VA_ARGS__)
490 491 492
#define kvm_pr_unimpl(fmt, ...) \
	pr_err_ratelimited("kvm [%i]: " fmt, \
			   task_tgid_nr(current), ## __VA_ARGS__)
493

494 495
/* The guest did something we don't support. */
#define vcpu_unimpl(vcpu, fmt, ...)					\
496 497
	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
498

499 500
#define vcpu_debug(vcpu, fmt, ...)					\
	kvm_debug("vcpu%i " fmt, (vcpu)->vcpu_id, ## __VA_ARGS__)
501 502 503
#define vcpu_debug_ratelimited(vcpu, fmt, ...)				\
	kvm_debug_ratelimited("vcpu%i " fmt, (vcpu)->vcpu_id,           \
			      ## __VA_ARGS__)
504 505
#define vcpu_err(vcpu, fmt, ...)					\
	kvm_err("vcpu%i " fmt, (vcpu)->vcpu_id, ## __VA_ARGS__)
506

507 508 509
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,
510 511
				      lockdep_is_held(&kvm->slots_lock) ||
				      !refcount_read(&kvm->users_count));
512 513
}

514 515
static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i)
{
516 517 518 519
	/* 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).
	 */
520 521 522 523 524
	smp_rmb();
	return kvm->vcpus[i];
}

#define kvm_for_each_vcpu(idx, vcpup, kvm) \
525 526 527 528
	for (idx = 0; \
	     idx < atomic_read(&kvm->online_vcpus) && \
	     (vcpup = kvm_get_vcpu(kvm, idx)) != NULL; \
	     idx++)
529

530 531
static inline struct kvm_vcpu *kvm_get_vcpu_by_id(struct kvm *kvm, int id)
{
532
	struct kvm_vcpu *vcpu = NULL;
533 534
	int i;

535
	if (id < 0)
536
		return NULL;
537 538
	if (id < KVM_MAX_VCPUS)
		vcpu = kvm_get_vcpu(kvm, id);
539 540
	if (vcpu && vcpu->vcpu_id == id)
		return vcpu;
541 542 543 544 545 546
	kvm_for_each_vcpu(i, vcpu, kvm)
		if (vcpu->vcpu_id == id)
			return vcpu;
	return NULL;
}

547 548 549 550 551 552 553 554 555 556 557
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();
}

558 559
#define kvm_for_each_memslot(memslot, slots)	\
	for (memslot = &slots->memslots[0];	\
560 561
	      memslot < slots->memslots + KVM_MEM_SLOTS_NUM && memslot->npages;\
		memslot++)
562

R
Rusty Russell 已提交
563 564 565
int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);

566
void vcpu_load(struct kvm_vcpu *vcpu);
567 568
void vcpu_put(struct kvm_vcpu *vcpu);

569
#ifdef __KVM_HAVE_IOAPIC
570
void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm);
571
void kvm_arch_post_irq_routing_update(struct kvm *kvm);
572
#else
573
static inline void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm)
574 575
{
}
576
static inline void kvm_arch_post_irq_routing_update(struct kvm *kvm)
577 578
{
}
579 580
#endif

581
#ifdef CONFIG_HAVE_KVM_IRQFD
582 583 584 585 586 587 588 589 590 591 592 593
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
594
int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
595
		  struct module *module);
596
void kvm_exit(void);
A
Avi Kivity 已提交
597

I
Izik Eidus 已提交
598 599 600
void kvm_get_kvm(struct kvm *kvm);
void kvm_put_kvm(struct kvm *kvm);

601
static inline struct kvm_memslots *__kvm_memslots(struct kvm *kvm, int as_id)
602
{
603
	return srcu_dereference_check(kvm->memslots[as_id], &kvm->srcu,
604 605
			lockdep_is_held(&kvm->slots_lock) ||
			!refcount_read(&kvm->users_count));
606 607
}

608 609 610 611 612
static inline struct kvm_memslots *kvm_memslots(struct kvm *kvm)
{
	return __kvm_memslots(kvm, 0);
}

613 614
static inline struct kvm_memslots *kvm_vcpu_memslots(struct kvm_vcpu *vcpu)
{
615 616 617
	int as_id = kvm_arch_vcpu_memslots_id(vcpu);

	return __kvm_memslots(vcpu->kvm, as_id);
618 619
}

620 621 622
static inline struct kvm_memory_slot *
id_to_memslot(struct kvm_memslots *slots, int id)
{
623 624
	int index = slots->id_to_index[id];
	struct kvm_memory_slot *slot;
625

626
	slot = &slots->memslots[index];
627

628 629
	WARN_ON(slot->id != id);
	return slot;
630 631
}

632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649
/*
 * 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,
};

650
int kvm_set_memory_region(struct kvm *kvm,
651
			  const struct kvm_userspace_memory_region *mem);
652
int __kvm_set_memory_region(struct kvm *kvm,
653
			    const struct kvm_userspace_memory_region *mem);
654
void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free,
655
			   struct kvm_memory_slot *dont);
656 657
int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot,
			    unsigned long npages);
658
void kvm_arch_memslots_updated(struct kvm *kvm, u64 gen);
659 660
int kvm_arch_prepare_memory_region(struct kvm *kvm,
				struct kvm_memory_slot *memslot,
661
				const struct kvm_userspace_memory_region *mem,
662
				enum kvm_mr_change change);
663
void kvm_arch_commit_memory_region(struct kvm *kvm,
664
				const struct kvm_userspace_memory_region *mem,
665
				const struct kvm_memory_slot *old,
666
				const struct kvm_memory_slot *new,
667
				enum kvm_mr_change change);
668
bool kvm_largepages_enabled(void);
669
void kvm_disable_largepages(void);
670 671 672 673 674
/* 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);
675

676 677
int gfn_to_page_many_atomic(struct kvm_memory_slot *slot, gfn_t gfn,
			    struct page **pages, int nr_pages);
678

A
Avi Kivity 已提交
679
struct page *gfn_to_page(struct kvm *kvm, gfn_t gfn);
M
Marcelo Tosatti 已提交
680
unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn);
681
unsigned long gfn_to_hva_prot(struct kvm *kvm, gfn_t gfn, bool *writable);
X
Xiao Guangrong 已提交
682
unsigned long gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
683 684
unsigned long gfn_to_hva_memslot_prot(struct kvm_memory_slot *slot, gfn_t gfn,
				      bool *writable);
685 686
void kvm_release_page_clean(struct page *page);
void kvm_release_page_dirty(struct page *page);
687 688
void kvm_set_page_accessed(struct page *page);

D
Dan Williams 已提交
689 690 691
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,
692
		      bool *writable);
D
Dan Williams 已提交
693 694 695 696 697
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);
698

D
Dan Williams 已提交
699
void kvm_release_pfn_clean(kvm_pfn_t pfn);
700
void kvm_release_pfn_dirty(kvm_pfn_t pfn);
D
Dan Williams 已提交
701 702 703
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);
704

705 706
int kvm_read_guest_page(struct kvm *kvm, gfn_t gfn, void *data, int offset,
			int len);
707 708
int kvm_read_guest_atomic(struct kvm *kvm, gpa_t gpa, void *data,
			  unsigned long len);
709
int kvm_read_guest(struct kvm *kvm, gpa_t gpa, void *data, unsigned long len);
710 711
int kvm_read_guest_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, unsigned long len);
712 713 714 715
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);
716 717 718
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,
719 720
				  void *data, unsigned int offset,
				  unsigned long len);
721 722
int kvm_gfn_to_hva_cache_init(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			      gpa_t gpa, unsigned long len);
723 724
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
725
struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn);
726
bool kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn);
J
Joerg Roedel 已提交
727
unsigned long kvm_host_page_size(struct kvm *kvm, gfn_t gfn);
A
Avi Kivity 已提交
728 729
void mark_page_dirty(struct kvm *kvm, gfn_t gfn);

730 731
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
732 733
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);
734 735 736 737 738 739 740 741 742 743 744 745 746 747 748
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);

749 750 751
void kvm_sigset_activate(struct kvm_vcpu *vcpu);
void kvm_sigset_deactivate(struct kvm_vcpu *vcpu);

752
void kvm_vcpu_block(struct kvm_vcpu *vcpu);
753 754
void kvm_arch_vcpu_blocking(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_unblocking(struct kvm_vcpu *vcpu);
755
bool kvm_vcpu_wake_up(struct kvm_vcpu *vcpu);
756
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
757
int kvm_vcpu_yield_to(struct kvm_vcpu *target);
758
void kvm_vcpu_on_spin(struct kvm_vcpu *vcpu, bool usermode_vcpu_not_eligible);
759

760
void kvm_flush_remote_tlbs(struct kvm *kvm);
761
void kvm_reload_remote_mmus(struct kvm *kvm);
762 763 764

bool kvm_make_vcpus_request_mask(struct kvm *kvm, unsigned int req,
				 unsigned long *vcpu_bitmap, cpumask_var_t tmp);
765
bool kvm_make_all_cpus_request(struct kvm *kvm, unsigned int req);
A
Avi Kivity 已提交
766

767 768
long kvm_arch_dev_ioctl(struct file *filp,
			unsigned int ioctl, unsigned long arg);
769 770
long kvm_arch_vcpu_ioctl(struct file *filp,
			 unsigned int ioctl, unsigned long arg);
771
vm_fault_t kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf);
772

773
int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext);
774

775 776
int kvm_get_dirty_log(struct kvm *kvm,
			struct kvm_dirty_log *log, int *is_dirty);
777 778

int kvm_get_dirty_log_protect(struct kvm *kvm,
779
			      struct kvm_dirty_log *log, bool *flush);
780 781
int kvm_clear_dirty_log_protect(struct kvm *kvm,
				struct kvm_clear_dirty_log *log, bool *flush);
782

783
void kvm_arch_mmu_enable_log_dirty_pt_masked(struct kvm *kvm,
784 785 786 787
					struct kvm_memory_slot *slot,
					gfn_t gfn_offset,
					unsigned long mask);

788 789
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
				struct kvm_dirty_log *log);
790 791
int kvm_vm_ioctl_clear_dirty_log(struct kvm *kvm,
				  struct kvm_clear_dirty_log *log);
792

793 794
int kvm_vm_ioctl_irq_line(struct kvm *kvm, struct kvm_irq_level *irq_level,
			bool line_status);
795 796
int kvm_vm_ioctl_enable_cap(struct kvm *kvm,
			    struct kvm_enable_cap *cap);
797 798
long kvm_arch_vm_ioctl(struct file *filp,
		       unsigned int ioctl, unsigned long arg);
799

800 801 802
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);

803 804 805
int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				    struct kvm_translation *tr);

806 807 808 809 810 811
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);
812 813 814 815
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
816 817
int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
					struct kvm_guest_debug *dbg);
818 819
int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run);

820 821
int kvm_arch_init(void *opaque);
void kvm_arch_exit(void);
822

823 824 825
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu);

R
Radim Krčmář 已提交
826 827
void kvm_arch_sched_in(struct kvm_vcpu *vcpu, int cpu);

828 829 830 831
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);
832
int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu);
833
void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu);
834
void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu);
835

836 837 838
bool kvm_arch_has_vcpu_debugfs(void);
int kvm_arch_create_vcpu_debugfs(struct kvm_vcpu *vcpu);

839 840
int kvm_arch_hardware_enable(void);
void kvm_arch_hardware_disable(void);
841 842 843
int kvm_arch_hardware_setup(void);
void kvm_arch_hardware_unsetup(void);
void kvm_arch_check_processor_compat(void *rtn);
844
int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu);
845
bool kvm_arch_vcpu_in_kernel(struct kvm_vcpu *vcpu);
846
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu);
847

848
#ifndef __KVM_HAVE_ARCH_VM_ALLOC
849 850 851 852
/*
 * All architectures that want to use vzalloc currently also
 * need their own kvm_arch_alloc_vm implementation.
 */
853 854 855 856 857 858 859 860 861 862 863
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

864 865 866 867 868 869 870
#ifndef __KVM_HAVE_ARCH_FLUSH_REMOTE_TLB
static inline int kvm_arch_flush_remote_tlb(struct kvm *kvm)
{
	return -ENOTSUPP;
}
#endif

871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888
#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
889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906
#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
907

908
static inline struct swait_queue_head *kvm_arch_vcpu_wq(struct kvm_vcpu *vcpu)
909
{
910 911 912
#ifdef __KVM_HAVE_ARCH_WQP
	return vcpu->arch.wqp;
#else
913 914
	return &vcpu->wq;
#endif
915
}
916

917 918 919 920 921 922 923 924 925 926 927 928 929 930
#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

931
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type);
932
void kvm_arch_destroy_vm(struct kvm *kvm);
933
void kvm_arch_sync_events(struct kvm *kvm);
934

935
int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu);
936
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
937

D
Dan Williams 已提交
938
bool kvm_is_reserved_pfn(kvm_pfn_t pfn);
939

B
Ben-Ami Yassour 已提交
940 941 942 943 944 945
struct kvm_irq_ack_notifier {
	struct hlist_node link;
	unsigned gsi;
	void (*irq_acked)(struct kvm_irq_ack_notifier *kian);
};

946 947 948
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);
949

950 951
int kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 irq, int level,
		bool line_status);
952
int kvm_set_msi(struct kvm_kernel_irq_routing_entry *irq_entry, struct kvm *kvm,
953
		int irq_source_id, int level, bool line_status);
954 955 956
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);
957
bool kvm_irq_has_notifier(struct kvm *kvm, unsigned irqchip, unsigned pin);
958
void kvm_notify_acked_gsi(struct kvm *kvm, int gsi);
959
void kvm_notify_acked_irq(struct kvm *kvm, unsigned irqchip, unsigned pin);
960 961
void kvm_register_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
962 963
void kvm_unregister_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
964 965
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
966

967 968 969 970 971 972 973 974 975
/*
 * 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)
{
976
	int start = 0, end = slots->used_slots;
977
	int slot = atomic_read(&slots->lru_slot);
978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997
	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];
	}
998 999 1000 1001 1002 1003 1004 1005 1006 1007

	return NULL;
}

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

1008 1009 1010 1011 1012 1013
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;
}

1014 1015 1016 1017 1018
static inline int memslot_id(struct kvm *kvm, gfn_t gfn)
{
	return gfn_to_memslot(kvm, gfn)->id;
}

1019 1020
static inline gfn_t
hva_to_gfn_memslot(unsigned long hva, struct kvm_memory_slot *slot)
1021
{
1022 1023 1024
	gfn_t gfn_offset = (hva - slot->userspace_addr) >> PAGE_SHIFT;

	return slot->base_gfn + gfn_offset;
1025 1026
}

A
Avi Kivity 已提交
1027 1028 1029 1030
static inline gpa_t gfn_to_gpa(gfn_t gfn)
{
	return (gpa_t)gfn << PAGE_SHIFT;
}
A
Avi Kivity 已提交
1031

1032 1033 1034 1035 1036
static inline gfn_t gpa_to_gfn(gpa_t gpa)
{
	return (gfn_t)(gpa >> PAGE_SHIFT);
}

D
Dan Williams 已提交
1037
static inline hpa_t pfn_to_hpa(kvm_pfn_t pfn)
B
Ben-Ami Yassour 已提交
1038 1039 1040 1041
{
	return (hpa_t)pfn << PAGE_SHIFT;
}

1042 1043 1044 1045 1046 1047
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));
}

1048 1049 1050 1051 1052 1053 1054
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);
}

1055 1056 1057 1058 1059
enum kvm_stat_kind {
	KVM_STAT_VM,
	KVM_STAT_VCPU,
};

1060 1061 1062 1063 1064
struct kvm_stat_data {
	int offset;
	struct kvm *kvm;
};

1065 1066 1067
struct kvm_stats_debugfs_item {
	const char *name;
	int offset;
1068
	enum kvm_stat_kind kind;
1069 1070
};
extern struct kvm_stats_debugfs_item debugfs_entries[];
1071
extern struct dentry *kvm_debugfs_dir;
1072

1073
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
1074
static inline int mmu_notifier_retry(struct kvm *kvm, unsigned long mmu_seq)
1075
{
1076
	if (unlikely(kvm->mmu_notifier_count))
1077 1078
		return 1;
	/*
1079 1080 1081 1082 1083 1084 1085 1086
	 * 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.
1087
	 */
1088
	smp_rmb();
1089
	if (kvm->mmu_notifier_seq != mmu_seq)
1090 1091 1092 1093 1094
		return 1;
	return 0;
}
#endif

1095
#ifdef CONFIG_HAVE_KVM_IRQ_ROUTING
1096

1097
#define KVM_MAX_IRQ_ROUTES 4096 /* might need extension/rework in the future */
1098

1099
bool kvm_arch_can_set_irq_routing(struct kvm *kvm);
1100 1101 1102 1103
int kvm_set_irq_routing(struct kvm *kvm,
			const struct kvm_irq_routing_entry *entries,
			unsigned nr,
			unsigned flags);
1104 1105
int kvm_set_routing_entry(struct kvm *kvm,
			  struct kvm_kernel_irq_routing_entry *e,
1106
			  const struct kvm_irq_routing_entry *ue);
1107 1108 1109 1110 1111 1112 1113 1114
void kvm_free_irq_routing(struct kvm *kvm);

#else

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

#endif

1115 1116
int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi);

G
Gregory Haskins 已提交
1117 1118
#ifdef CONFIG_HAVE_KVM_EVENTFD

G
Gregory Haskins 已提交
1119
void kvm_eventfd_init(struct kvm *kvm);
1120 1121
int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args);

1122
#ifdef CONFIG_HAVE_KVM_IRQFD
1123
int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args);
G
Gregory Haskins 已提交
1124
void kvm_irqfd_release(struct kvm *kvm);
1125
void kvm_irq_routing_update(struct kvm *);
1126 1127 1128 1129 1130 1131 1132 1133
#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
1134 1135 1136

#else

G
Gregory Haskins 已提交
1137
static inline void kvm_eventfd_init(struct kvm *kvm) {}
1138

1139
static inline int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args)
G
Gregory Haskins 已提交
1140 1141 1142 1143 1144
{
	return -EINVAL;
}

static inline void kvm_irqfd_release(struct kvm *kvm) {}
1145

A
Alexander Graf 已提交
1146
#ifdef CONFIG_HAVE_KVM_IRQCHIP
1147
static inline void kvm_irq_routing_update(struct kvm *kvm)
1148 1149
{
}
A
Alexander Graf 已提交
1150
#endif
1151

G
Gregory Haskins 已提交
1152 1153 1154 1155
static inline int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args)
{
	return -ENOSYS;
}
G
Gregory Haskins 已提交
1156 1157 1158

#endif /* CONFIG_HAVE_KVM_EVENTFD */

1159 1160
void kvm_arch_irq_routing_update(struct kvm *kvm);

1161 1162
static inline void kvm_make_request(int req, struct kvm_vcpu *vcpu)
{
1163 1164 1165 1166 1167
	/*
	 * 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();
1168
	set_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests);
1169 1170
}

R
Radim Krčmář 已提交
1171 1172 1173 1174 1175
static inline bool kvm_request_pending(struct kvm_vcpu *vcpu)
{
	return READ_ONCE(vcpu->requests);
}

1176 1177
static inline bool kvm_test_request(int req, struct kvm_vcpu *vcpu)
{
1178
	return test_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests);
1179 1180 1181 1182
}

static inline void kvm_clear_request(int req, struct kvm_vcpu *vcpu)
{
1183
	clear_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests);
1184 1185
}

1186 1187
static inline bool kvm_check_request(int req, struct kvm_vcpu *vcpu)
{
1188 1189
	if (kvm_test_request(req, vcpu)) {
		kvm_clear_request(req, vcpu);
1190 1191 1192 1193 1194 1195

		/*
		 * 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();
1196 1197 1198 1199
		return true;
	} else {
		return false;
	}
1200 1201
}

1202 1203
extern bool kvm_rebooting;

1204 1205
extern unsigned int halt_poll_ns;
extern unsigned int halt_poll_ns_grow;
1206
extern unsigned int halt_poll_ns_grow_start;
1207 1208
extern unsigned int halt_poll_ns_shrink;

S
Scott Wood 已提交
1209 1210 1211 1212
struct kvm_device {
	struct kvm_device_ops *ops;
	struct kvm *kvm;
	void *private;
1213
	struct list_head vm_node;
S
Scott Wood 已提交
1214 1215 1216 1217 1218
};

/* create, destroy, and name are mandatory */
struct kvm_device_ops {
	const char *name;
1219 1220 1221 1222 1223 1224

	/*
	 * 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
1225 1226
	int (*create)(struct kvm_device *dev, u32 type);

1227 1228 1229 1230 1231 1232
	/*
	 * 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
1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252
	/*
	 * 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);
1253
int kvm_register_device_ops(struct kvm_device_ops *ops, u32 type);
1254
void kvm_unregister_device_ops(u32 type);
S
Scott Wood 已提交
1255

1256
extern struct kvm_device_ops kvm_mpic_ops;
1257
extern struct kvm_device_ops kvm_arm_vgic_v2_ops;
1258
extern struct kvm_device_ops kvm_arm_vgic_v3_ops;
1259

1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280
#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 */
1281 1282

#ifdef CONFIG_HAVE_KVM_IRQ_BYPASS
1283
bool kvm_arch_has_irq_bypass(void);
1284 1285 1286 1287 1288 1289
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 *);
1290 1291
int kvm_arch_update_irqfd_routing(struct kvm *kvm, unsigned int host_irq,
				  uint32_t guest_irq, bool set);
1292
#endif /* CONFIG_HAVE_KVM_IRQ_BYPASS */
1293

1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307
#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 */

1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319
#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 */

1320 1321
int kvm_arch_mmu_notifier_invalidate_range(struct kvm *kvm,
		unsigned long start, unsigned long end, bool blockable);
1322

1323 1324 1325 1326 1327 1328 1329 1330 1331
#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 */

1332
#endif