svm.c 142.2 KB
Newer Older
A
Avi Kivity 已提交
1 2 3 4 5 6
/*
 * Kernel-based Virtual Machine driver for Linux
 *
 * AMD SVM support
 *
 * Copyright (C) 2006 Qumranet, Inc.
N
Nicolas Kaiser 已提交
7
 * Copyright 2010 Red Hat, Inc. and/or its affiliates.
A
Avi Kivity 已提交
8 9 10 11 12 13 14 15 16
 *
 * Authors:
 *   Yaniv Kamay  <yaniv@qumranet.com>
 *   Avi Kivity   <avi@qumranet.com>
 *
 * This work is licensed under the terms of the GNU GPL, version 2.  See
 * the COPYING file in the top-level directory.
 *
 */
17 18 19

#define pr_fmt(fmt) "SVM: " fmt

20 21
#include <linux/kvm_host.h>

22
#include "irq.h"
23
#include "mmu.h"
24
#include "kvm_cache_regs.h"
25
#include "x86.h"
26
#include "cpuid.h"
27
#include "pmu.h"
A
Avi Kivity 已提交
28

A
Avi Kivity 已提交
29
#include <linux/module.h>
30
#include <linux/mod_devicetable.h>
31
#include <linux/kernel.h>
A
Avi Kivity 已提交
32 33
#include <linux/vmalloc.h>
#include <linux/highmem.h>
A
Alexey Dobriyan 已提交
34
#include <linux/sched.h>
35
#include <linux/trace_events.h>
36
#include <linux/slab.h>
37 38
#include <linux/amd-iommu.h>
#include <linux/hashtable.h>
39
#include <linux/frame.h>
A
Avi Kivity 已提交
40

41
#include <asm/apic.h>
42
#include <asm/perf_event.h>
43
#include <asm/tlbflush.h>
A
Avi Kivity 已提交
44
#include <asm/desc.h>
45
#include <asm/debugreg.h>
G
Gleb Natapov 已提交
46
#include <asm/kvm_para.h>
47
#include <asm/irq_remapping.h>
A
Avi Kivity 已提交
48

49
#include <asm/virtext.h>
50
#include "trace.h"
51

52 53
#define __ex(x) __kvm_handle_fault_on_reboot(x)

A
Avi Kivity 已提交
54 55 56
MODULE_AUTHOR("Qumranet");
MODULE_LICENSE("GPL");

57 58 59 60 61 62
static const struct x86_cpu_id svm_cpu_id[] = {
	X86_FEATURE_MATCH(X86_FEATURE_SVM),
	{}
};
MODULE_DEVICE_TABLE(x86cpu, svm_cpu_id);

A
Avi Kivity 已提交
63 64 65 66 67 68
#define IOPM_ALLOC_ORDER 2
#define MSRPM_ALLOC_ORDER 1

#define SEG_TYPE_LDT 2
#define SEG_TYPE_BUSY_TSS16 3

69 70 71 72
#define SVM_FEATURE_NPT            (1 <<  0)
#define SVM_FEATURE_LBRV           (1 <<  1)
#define SVM_FEATURE_SVML           (1 <<  2)
#define SVM_FEATURE_NRIP           (1 <<  3)
73 74 75 76
#define SVM_FEATURE_TSC_RATE       (1 <<  4)
#define SVM_FEATURE_VMCB_CLEAN     (1 <<  5)
#define SVM_FEATURE_FLUSH_ASID     (1 <<  6)
#define SVM_FEATURE_DECODE_ASSIST  (1 <<  7)
77
#define SVM_FEATURE_PAUSE_FILTER   (1 << 10)
78

79 80
#define SVM_AVIC_DOORBELL	0xc001011b

81 82 83 84
#define NESTED_EXIT_HOST	0	/* Exit handled on host level */
#define NESTED_EXIT_DONE	1	/* Exit caused nested vmexit  */
#define NESTED_EXIT_CONTINUE	2	/* Further checks needed      */

85 86
#define DEBUGCTL_RESERVED_BITS (~(0x3fULL))

87
#define TSC_RATIO_RSVD          0xffffff0000000000ULL
88 89
#define TSC_RATIO_MIN		0x0000000000000001ULL
#define TSC_RATIO_MAX		0x000000ffffffffffULL
90

D
Dan Carpenter 已提交
91
#define AVIC_HPA_MASK	~((0xFFFULL << 52) | 0xFFF)
92 93 94 95 96 97 98

/*
 * 0xff is broadcast, so the max index allowed for physical APIC ID
 * table is 0xfe.  APIC IDs above 0xff are reserved.
 */
#define AVIC_MAX_PHYSICAL_ID_COUNT	255

99 100 101 102
#define AVIC_UNACCEL_ACCESS_WRITE_MASK		1
#define AVIC_UNACCEL_ACCESS_OFFSET_MASK		0xFF0
#define AVIC_UNACCEL_ACCESS_VECTOR_MASK		0xFFFFFFFF

103 104 105 106 107 108 109 110 111 112 113 114 115
/* AVIC GATAG is encoded using VM and VCPU IDs */
#define AVIC_VCPU_ID_BITS		8
#define AVIC_VCPU_ID_MASK		((1 << AVIC_VCPU_ID_BITS) - 1)

#define AVIC_VM_ID_BITS			24
#define AVIC_VM_ID_NR			(1 << AVIC_VM_ID_BITS)
#define AVIC_VM_ID_MASK			((1 << AVIC_VM_ID_BITS) - 1)

#define AVIC_GATAG(x, y)		(((x & AVIC_VM_ID_MASK) << AVIC_VCPU_ID_BITS) | \
						(y & AVIC_VCPU_ID_MASK))
#define AVIC_GATAG_TO_VMID(x)		((x >> AVIC_VCPU_ID_BITS) & AVIC_VM_ID_MASK)
#define AVIC_GATAG_TO_VCPUID(x)		(x & AVIC_VCPU_ID_MASK)

116 117
static bool erratum_383_found __read_mostly;

A
Avi Kivity 已提交
118 119 120 121 122 123
static const u32 host_save_user_msrs[] = {
#ifdef CONFIG_X86_64
	MSR_STAR, MSR_LSTAR, MSR_CSTAR, MSR_SYSCALL_MASK, MSR_KERNEL_GS_BASE,
	MSR_FS_BASE,
#endif
	MSR_IA32_SYSENTER_CS, MSR_IA32_SYSENTER_ESP, MSR_IA32_SYSENTER_EIP,
P
Paolo Bonzini 已提交
124
	MSR_TSC_AUX,
A
Avi Kivity 已提交
125 126 127 128 129 130
};

#define NR_HOST_SAVE_USER_MSRS ARRAY_SIZE(host_save_user_msrs)

struct kvm_vcpu;

131 132 133
struct nested_state {
	struct vmcb *hsave;
	u64 hsave_msr;
134
	u64 vm_cr_msr;
135 136 137 138 139 140 141
	u64 vmcb;

	/* These are the merged vectors */
	u32 *msrpm;

	/* gpa pointers to the real vectors */
	u64 vmcb_msrpm;
142
	u64 vmcb_iopm;
J
Joerg Roedel 已提交
143

144 145 146
	/* A VMEXIT is required but not yet emulated */
	bool exit_required;

J
Joerg Roedel 已提交
147
	/* cache for intercepts of the guest */
148
	u32 intercept_cr;
149
	u32 intercept_dr;
J
Joerg Roedel 已提交
150 151 152
	u32 intercept_exceptions;
	u64 intercept;

153 154
	/* Nested Paging related state */
	u64 nested_cr3;
155 156
};

157 158 159
#define MSRPM_OFFSETS	16
static u32 msrpm_offsets[MSRPM_OFFSETS] __read_mostly;

160 161 162 163 164 165
/*
 * Set osvw_len to higher value when updated Revision Guides
 * are published and we know what the new status bits are
 */
static uint64_t osvw_len = 4, osvw_status;

A
Avi Kivity 已提交
166 167 168 169 170 171 172 173
struct vcpu_svm {
	struct kvm_vcpu vcpu;
	struct vmcb *vmcb;
	unsigned long vmcb_pa;
	struct svm_cpu_data *svm_data;
	uint64_t asid_generation;
	uint64_t sysenter_esp;
	uint64_t sysenter_eip;
P
Paolo Bonzini 已提交
174
	uint64_t tsc_aux;
A
Avi Kivity 已提交
175 176 177 178

	u64 next_rip;

	u64 host_user_msrs[NR_HOST_SAVE_USER_MSRS];
179
	struct {
180 181 182
		u16 fs;
		u16 gs;
		u16 ldt;
183 184
		u64 gs_base;
	} host;
A
Avi Kivity 已提交
185 186 187

	u32 *msrpm;

188 189
	ulong nmi_iret_rip;

190
	struct nested_state nested;
J
Jan Kiszka 已提交
191 192

	bool nmi_singlestep;
193
	u64 nmi_singlestep_guest_rflags;
194 195 196

	unsigned int3_injected;
	unsigned long int3_rip;
197

198 199
	/* cached guest cpuid flags for faster access */
	bool nrips_enabled	: 1;
200

201
	u32 ldr_reg;
202 203
	struct page *avic_backing_page;
	u64 *avic_physical_id_cache;
204
	bool avic_is_running;
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221

	/*
	 * Per-vcpu list of struct amd_svm_iommu_ir:
	 * This is used mainly to store interrupt remapping information used
	 * when update the vcpu affinity. This avoids the need to scan for
	 * IRTE and try to match ga_tag in the IOMMU driver.
	 */
	struct list_head ir_list;
	spinlock_t ir_list_lock;
};

/*
 * This is a wrapper of struct amd_iommu_ir_data.
 */
struct amd_svm_iommu_ir {
	struct list_head node;	/* Used by SVM for per-vcpu ir_list */
	void *data;		/* Storing pointer to struct amd_ir_data */
A
Avi Kivity 已提交
222 223
};

224 225 226 227 228 229 230 231
#define AVIC_LOGICAL_ID_ENTRY_GUEST_PHYSICAL_ID_MASK	(0xFF)
#define AVIC_LOGICAL_ID_ENTRY_VALID_MASK		(1 << 31)

#define AVIC_PHYSICAL_ID_ENTRY_HOST_PHYSICAL_ID_MASK	(0xFFULL)
#define AVIC_PHYSICAL_ID_ENTRY_BACKING_PAGE_MASK	(0xFFFFFFFFFFULL << 12)
#define AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK		(1ULL << 62)
#define AVIC_PHYSICAL_ID_ENTRY_VALID_MASK		(1ULL << 63)

232 233 234
static DEFINE_PER_CPU(u64, current_tsc_ratio);
#define TSC_RATIO_DEFAULT	0x0100000000ULL

235 236
#define MSR_INVALID			0xffffffffU

237
static const struct svm_direct_access_msrs {
238 239 240
	u32 index;   /* Index of the MSR */
	bool always; /* True if intercept is always on */
} direct_access_msrs[] = {
B
Brian Gerst 已提交
241
	{ .index = MSR_STAR,				.always = true  },
242 243 244 245 246 247 248 249 250 251 252 253 254 255
	{ .index = MSR_IA32_SYSENTER_CS,		.always = true  },
#ifdef CONFIG_X86_64
	{ .index = MSR_GS_BASE,				.always = true  },
	{ .index = MSR_FS_BASE,				.always = true  },
	{ .index = MSR_KERNEL_GS_BASE,			.always = true  },
	{ .index = MSR_LSTAR,				.always = true  },
	{ .index = MSR_CSTAR,				.always = true  },
	{ .index = MSR_SYSCALL_MASK,			.always = true  },
#endif
	{ .index = MSR_IA32_LASTBRANCHFROMIP,		.always = false },
	{ .index = MSR_IA32_LASTBRANCHTOIP,		.always = false },
	{ .index = MSR_IA32_LASTINTFROMIP,		.always = false },
	{ .index = MSR_IA32_LASTINTTOIP,		.always = false },
	{ .index = MSR_INVALID,				.always = false },
A
Avi Kivity 已提交
256 257
};

258 259 260 261
/* enable NPT for AMD64 and X86 with PAE */
#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE)
static bool npt_enabled = true;
#else
J
Joerg Roedel 已提交
262
static bool npt_enabled;
263
#endif
264

265 266
/* allow nested paging (virtualized MMU) for all guests */
static int npt = true;
267
module_param(npt, int, S_IRUGO);
268

269 270
/* allow nested virtualization in KVM/SVM */
static int nested = true;
271 272
module_param(nested, int, S_IRUGO);

273 274
/* enable / disable AVIC */
static int avic;
275
#ifdef CONFIG_X86_LOCAL_APIC
276
module_param(avic, int, S_IRUGO);
277
#endif
278

279 280 281 282
/* enable/disable Virtual VMLOAD VMSAVE */
static int vls = true;
module_param(vls, int, 0444);

283 284 285
/* enable/disable Virtual GIF */
static int vgif = true;
module_param(vgif, int, 0444);
286

287
static void svm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0);
288
static void svm_flush_tlb(struct kvm_vcpu *vcpu);
289
static void svm_complete_interrupts(struct vcpu_svm *svm);
290

291
static int nested_svm_exit_handled(struct vcpu_svm *svm);
292
static int nested_svm_intercept(struct vcpu_svm *svm);
293 294 295 296
static int nested_svm_vmexit(struct vcpu_svm *svm);
static int nested_svm_check_exception(struct vcpu_svm *svm, unsigned nr,
				      bool has_error_code, u32 error_code);

297
enum {
298 299
	VMCB_INTERCEPTS, /* Intercept vectors, TSC offset,
			    pause filter count */
300
	VMCB_PERM_MAP,   /* IOPM Base and MSRPM Base */
301
	VMCB_ASID,	 /* ASID */
302
	VMCB_INTR,	 /* int_ctl, int_vector */
303
	VMCB_NPT,        /* npt_en, nCR3, gPAT */
304
	VMCB_CR,	 /* CR0, CR3, CR4, EFER */
305
	VMCB_DR,         /* DR6, DR7 */
306
	VMCB_DT,         /* GDT, IDT */
307
	VMCB_SEG,        /* CS, DS, SS, ES, CPL */
308
	VMCB_CR2,        /* CR2 only */
309
	VMCB_LBR,        /* DBGCTL, BR_FROM, BR_TO, LAST_EX_FROM, LAST_EX_TO */
310 311 312 313
	VMCB_AVIC,       /* AVIC APIC_BAR, AVIC APIC_BACKING_PAGE,
			  * AVIC PHYSICAL_TABLE pointer,
			  * AVIC LOGICAL_TABLE pointer
			  */
314 315 316
	VMCB_DIRTY_MAX,
};

317 318
/* TPR and CR2 are always written before VMRUN */
#define VMCB_ALWAYS_DIRTY_MASK	((1U << VMCB_INTR) | (1U << VMCB_CR2))
319

320 321
#define VMCB_AVIC_APIC_BAR_MASK		0xFFFFFFFFFF000ULL

322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337
static inline void mark_all_dirty(struct vmcb *vmcb)
{
	vmcb->control.clean = 0;
}

static inline void mark_all_clean(struct vmcb *vmcb)
{
	vmcb->control.clean = ((1 << VMCB_DIRTY_MAX) - 1)
			       & ~VMCB_ALWAYS_DIRTY_MASK;
}

static inline void mark_dirty(struct vmcb *vmcb, int bit)
{
	vmcb->control.clean &= ~(1 << bit);
}

338 339
static inline struct vcpu_svm *to_svm(struct kvm_vcpu *vcpu)
{
R
Rusty Russell 已提交
340
	return container_of(vcpu, struct vcpu_svm, vcpu);
341 342
}

343 344 345 346 347 348
static inline void avic_update_vapic_bar(struct vcpu_svm *svm, u64 data)
{
	svm->vmcb->control.avic_vapic_bar = data & VMCB_AVIC_APIC_BAR_MASK;
	mark_dirty(svm->vmcb, VMCB_AVIC);
}

349 350 351 352 353 354 355 356 357 358 359
static inline bool avic_vcpu_is_running(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u64 *entry = svm->avic_physical_id_cache;

	if (!entry)
		return false;

	return (READ_ONCE(*entry) & AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK);
}

360 361 362 363 364
static void recalc_intercepts(struct vcpu_svm *svm)
{
	struct vmcb_control_area *c, *h;
	struct nested_state *g;

365 366
	mark_dirty(svm->vmcb, VMCB_INTERCEPTS);

367 368 369 370 371 372 373
	if (!is_guest_mode(&svm->vcpu))
		return;

	c = &svm->vmcb->control;
	h = &svm->nested.hsave->control;
	g = &svm->nested;

374
	c->intercept_cr = h->intercept_cr | g->intercept_cr;
375
	c->intercept_dr = h->intercept_dr | g->intercept_dr;
376 377 378 379
	c->intercept_exceptions = h->intercept_exceptions | g->intercept_exceptions;
	c->intercept = h->intercept | g->intercept;
}

380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412
static inline struct vmcb *get_host_vmcb(struct vcpu_svm *svm)
{
	if (is_guest_mode(&svm->vcpu))
		return svm->nested.hsave;
	else
		return svm->vmcb;
}

static inline void set_cr_intercept(struct vcpu_svm *svm, int bit)
{
	struct vmcb *vmcb = get_host_vmcb(svm);

	vmcb->control.intercept_cr |= (1U << bit);

	recalc_intercepts(svm);
}

static inline void clr_cr_intercept(struct vcpu_svm *svm, int bit)
{
	struct vmcb *vmcb = get_host_vmcb(svm);

	vmcb->control.intercept_cr &= ~(1U << bit);

	recalc_intercepts(svm);
}

static inline bool is_cr_intercept(struct vcpu_svm *svm, int bit)
{
	struct vmcb *vmcb = get_host_vmcb(svm);

	return vmcb->control.intercept_cr & (1U << bit);
}

413
static inline void set_dr_intercepts(struct vcpu_svm *svm)
414 415 416
{
	struct vmcb *vmcb = get_host_vmcb(svm);

417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
	vmcb->control.intercept_dr = (1 << INTERCEPT_DR0_READ)
		| (1 << INTERCEPT_DR1_READ)
		| (1 << INTERCEPT_DR2_READ)
		| (1 << INTERCEPT_DR3_READ)
		| (1 << INTERCEPT_DR4_READ)
		| (1 << INTERCEPT_DR5_READ)
		| (1 << INTERCEPT_DR6_READ)
		| (1 << INTERCEPT_DR7_READ)
		| (1 << INTERCEPT_DR0_WRITE)
		| (1 << INTERCEPT_DR1_WRITE)
		| (1 << INTERCEPT_DR2_WRITE)
		| (1 << INTERCEPT_DR3_WRITE)
		| (1 << INTERCEPT_DR4_WRITE)
		| (1 << INTERCEPT_DR5_WRITE)
		| (1 << INTERCEPT_DR6_WRITE)
		| (1 << INTERCEPT_DR7_WRITE);
433 434 435 436

	recalc_intercepts(svm);
}

437
static inline void clr_dr_intercepts(struct vcpu_svm *svm)
438 439 440
{
	struct vmcb *vmcb = get_host_vmcb(svm);

441
	vmcb->control.intercept_dr = 0;
442 443 444 445

	recalc_intercepts(svm);
}

446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463
static inline void set_exception_intercept(struct vcpu_svm *svm, int bit)
{
	struct vmcb *vmcb = get_host_vmcb(svm);

	vmcb->control.intercept_exceptions |= (1U << bit);

	recalc_intercepts(svm);
}

static inline void clr_exception_intercept(struct vcpu_svm *svm, int bit)
{
	struct vmcb *vmcb = get_host_vmcb(svm);

	vmcb->control.intercept_exceptions &= ~(1U << bit);

	recalc_intercepts(svm);
}

464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481
static inline void set_intercept(struct vcpu_svm *svm, int bit)
{
	struct vmcb *vmcb = get_host_vmcb(svm);

	vmcb->control.intercept |= (1ULL << bit);

	recalc_intercepts(svm);
}

static inline void clr_intercept(struct vcpu_svm *svm, int bit)
{
	struct vmcb *vmcb = get_host_vmcb(svm);

	vmcb->control.intercept &= ~(1ULL << bit);

	recalc_intercepts(svm);
}

482 483 484 485 486
static inline bool vgif_enabled(struct vcpu_svm *svm)
{
	return !!(svm->vmcb->control.int_ctl & V_GIF_ENABLE_MASK);
}

487 488
static inline void enable_gif(struct vcpu_svm *svm)
{
489 490 491 492
	if (vgif_enabled(svm))
		svm->vmcb->control.int_ctl |= V_GIF_MASK;
	else
		svm->vcpu.arch.hflags |= HF_GIF_MASK;
493 494 495 496
}

static inline void disable_gif(struct vcpu_svm *svm)
{
497 498 499 500
	if (vgif_enabled(svm))
		svm->vmcb->control.int_ctl &= ~V_GIF_MASK;
	else
		svm->vcpu.arch.hflags &= ~HF_GIF_MASK;
501 502 503 504
}

static inline bool gif_set(struct vcpu_svm *svm)
{
505 506 507 508
	if (vgif_enabled(svm))
		return !!(svm->vmcb->control.int_ctl & V_GIF_MASK);
	else
		return !!(svm->vcpu.arch.hflags & HF_GIF_MASK);
509 510
}

511
static unsigned long iopm_base;
A
Avi Kivity 已提交
512 513 514 515

struct kvm_ldttss_desc {
	u16 limit0;
	u16 base0;
J
Joerg Roedel 已提交
516 517
	unsigned base1:8, type:5, dpl:2, p:1;
	unsigned limit1:4, zero0:3, g:1, base2:8;
A
Avi Kivity 已提交
518 519 520 521 522 523 524
	u32 base3;
	u32 zero1;
} __attribute__((packed));

struct svm_cpu_data {
	int cpu;

A
Avi Kivity 已提交
525 526 527
	u64 asid_generation;
	u32 max_asid;
	u32 next_asid;
A
Avi Kivity 已提交
528 529 530 531 532 533 534 535 536 537 538 539
	struct kvm_ldttss_desc *tss_desc;

	struct page *save_area;
};

static DEFINE_PER_CPU(struct svm_cpu_data *, svm_data);

struct svm_init_data {
	int cpu;
	int r;
};

540
static const u32 msrpm_ranges[] = {0, 0xc0000000, 0xc0010000};
A
Avi Kivity 已提交
541

542
#define NUM_MSR_MAPS ARRAY_SIZE(msrpm_ranges)
A
Avi Kivity 已提交
543 544 545
#define MSRS_RANGE_SIZE 2048
#define MSRS_IN_RANGE (MSRS_RANGE_SIZE * 8 / 2)

546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566
static u32 svm_msrpm_offset(u32 msr)
{
	u32 offset;
	int i;

	for (i = 0; i < NUM_MSR_MAPS; i++) {
		if (msr < msrpm_ranges[i] ||
		    msr >= msrpm_ranges[i] + MSRS_IN_RANGE)
			continue;

		offset  = (msr - msrpm_ranges[i]) / 4; /* 4 msrs per u8 */
		offset += (i * MSRS_RANGE_SIZE);       /* add range offset */

		/* Now we have the u8 offset - but need the u32 offset */
		return offset / 4;
	}

	/* MSR not in any range */
	return MSR_INVALID;
}

A
Avi Kivity 已提交
567 568 569 570
#define MAX_INST_SIZE 15

static inline void clgi(void)
{
571
	asm volatile (__ex(SVM_CLGI));
A
Avi Kivity 已提交
572 573 574 575
}

static inline void stgi(void)
{
576
	asm volatile (__ex(SVM_STGI));
A
Avi Kivity 已提交
577 578 579 580
}

static inline void invlpga(unsigned long addr, u32 asid)
{
J
Joerg Roedel 已提交
581
	asm volatile (__ex(SVM_INVLPGA) : : "a"(addr), "c"(asid));
A
Avi Kivity 已提交
582 583
}

584
static int get_npt_level(struct kvm_vcpu *vcpu)
585 586
{
#ifdef CONFIG_X86_64
587
	return PT64_ROOT_4LEVEL;
588 589 590 591 592
#else
	return PT32E_ROOT_LEVEL;
#endif
}

A
Avi Kivity 已提交
593 594
static void svm_set_efer(struct kvm_vcpu *vcpu, u64 efer)
{
595
	vcpu->arch.efer = efer;
596
	if (!npt_enabled && !(efer & EFER_LMA))
597
		efer &= ~EFER_LME;
A
Avi Kivity 已提交
598

599
	to_svm(vcpu)->vmcb->save.efer = efer | EFER_SVME;
600
	mark_dirty(to_svm(vcpu)->vmcb, VMCB_CR);
A
Avi Kivity 已提交
601 602 603 604 605 606 607 608
}

static int is_external_interrupt(u32 info)
{
	info &= SVM_EVTINJ_TYPE_MASK | SVM_EVTINJ_VALID;
	return info == (SVM_EVTINJ_VALID | SVM_EVTINJ_TYPE_INTR);
}

609
static u32 svm_get_interrupt_shadow(struct kvm_vcpu *vcpu)
610 611 612 613 614
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u32 ret = 0;

	if (svm->vmcb->control.int_state & SVM_INTERRUPT_SHADOW_MASK)
615 616
		ret = KVM_X86_SHADOW_INT_STI | KVM_X86_SHADOW_INT_MOV_SS;
	return ret;
617 618 619 620 621 622 623 624 625 626 627 628 629
}

static void svm_set_interrupt_shadow(struct kvm_vcpu *vcpu, int mask)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	if (mask == 0)
		svm->vmcb->control.int_state &= ~SVM_INTERRUPT_SHADOW_MASK;
	else
		svm->vmcb->control.int_state |= SVM_INTERRUPT_SHADOW_MASK;

}

A
Avi Kivity 已提交
630 631
static void skip_emulated_instruction(struct kvm_vcpu *vcpu)
{
632 633
	struct vcpu_svm *svm = to_svm(vcpu);

634
	if (svm->vmcb->control.next_rip != 0) {
635
		WARN_ON_ONCE(!static_cpu_has(X86_FEATURE_NRIPS));
636
		svm->next_rip = svm->vmcb->control.next_rip;
637
	}
638

639
	if (!svm->next_rip) {
640
		if (emulate_instruction(vcpu, EMULTYPE_SKIP) !=
641 642
				EMULATE_DONE)
			printk(KERN_DEBUG "%s: NOP\n", __func__);
A
Avi Kivity 已提交
643 644
		return;
	}
645 646 647
	if (svm->next_rip - kvm_rip_read(vcpu) > MAX_INST_SIZE)
		printk(KERN_ERR "%s: ip 0x%lx next 0x%llx\n",
		       __func__, kvm_rip_read(vcpu), svm->next_rip);
A
Avi Kivity 已提交
648

649
	kvm_rip_write(vcpu, svm->next_rip);
650
	svm_set_interrupt_shadow(vcpu, 0);
A
Avi Kivity 已提交
651 652
}

653
static void svm_queue_exception(struct kvm_vcpu *vcpu)
J
Jan Kiszka 已提交
654 655
{
	struct vcpu_svm *svm = to_svm(vcpu);
656 657
	unsigned nr = vcpu->arch.exception.nr;
	bool has_error_code = vcpu->arch.exception.has_error_code;
658
	bool reinject = vcpu->arch.exception.injected;
659
	u32 error_code = vcpu->arch.exception.error_code;
J
Jan Kiszka 已提交
660

J
Joerg Roedel 已提交
661 662 663 664
	/*
	 * If we are within a nested VM we'd better #VMEXIT and let the guest
	 * handle the exception
	 */
665 666
	if (!reinject &&
	    nested_svm_check_exception(svm, nr, has_error_code, error_code))
J
Jan Kiszka 已提交
667 668
		return;

669
	if (nr == BP_VECTOR && !static_cpu_has(X86_FEATURE_NRIPS)) {
670 671 672 673 674 675 676 677 678 679 680 681 682 683 684
		unsigned long rip, old_rip = kvm_rip_read(&svm->vcpu);

		/*
		 * For guest debugging where we have to reinject #BP if some
		 * INT3 is guest-owned:
		 * Emulate nRIP by moving RIP forward. Will fail if injection
		 * raises a fault that is not intercepted. Still better than
		 * failing in all cases.
		 */
		skip_emulated_instruction(&svm->vcpu);
		rip = kvm_rip_read(&svm->vcpu);
		svm->int3_rip = rip + svm->vmcb->save.cs.base;
		svm->int3_injected = rip - old_rip;
	}

J
Jan Kiszka 已提交
685 686 687 688 689 690 691
	svm->vmcb->control.event_inj = nr
		| SVM_EVTINJ_VALID
		| (has_error_code ? SVM_EVTINJ_VALID_ERR : 0)
		| SVM_EVTINJ_TYPE_EXEPT;
	svm->vmcb->control.event_inj_err = error_code;
}

692 693 694 695 696 697
static void svm_init_erratum_383(void)
{
	u32 low, high;
	int err;
	u64 val;

698
	if (!static_cpu_has_bug(X86_BUG_AMD_TLB_MMATCH))
699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
		return;

	/* Use _safe variants to not break nested virtualization */
	val = native_read_msr_safe(MSR_AMD64_DC_CFG, &err);
	if (err)
		return;

	val |= (1ULL << 47);

	low  = lower_32_bits(val);
	high = upper_32_bits(val);

	native_write_msr_safe(MSR_AMD64_DC_CFG, low, high);

	erratum_383_found = true;
}

716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736
static void svm_init_osvw(struct kvm_vcpu *vcpu)
{
	/*
	 * Guests should see errata 400 and 415 as fixed (assuming that
	 * HLT and IO instructions are intercepted).
	 */
	vcpu->arch.osvw.length = (osvw_len >= 3) ? (osvw_len) : 3;
	vcpu->arch.osvw.status = osvw_status & ~(6ULL);

	/*
	 * By increasing VCPU's osvw.length to 3 we are telling the guest that
	 * all osvw.status bits inside that length, including bit 0 (which is
	 * reserved for erratum 298), are valid. However, if host processor's
	 * osvw_len is 0 then osvw_status[0] carries no information. We need to
	 * be conservative here and therefore we tell the guest that erratum 298
	 * is present (because we really don't know).
	 */
	if (osvw_len == 0 && boot_cpu_data.x86 == 0x10)
		vcpu->arch.osvw.status |= 1;
}

A
Avi Kivity 已提交
737 738
static int has_svm(void)
{
739
	const char *msg;
A
Avi Kivity 已提交
740

741
	if (!cpu_has_svm(&msg)) {
J
Joe Perches 已提交
742
		printk(KERN_INFO "has_svm: %s\n", msg);
A
Avi Kivity 已提交
743 744 745 746 747 748
		return 0;
	}

	return 1;
}

749
static void svm_hardware_disable(void)
A
Avi Kivity 已提交
750
{
751 752 753 754
	/* Make sure we clean up behind us */
	if (static_cpu_has(X86_FEATURE_TSCRATEMSR))
		wrmsrl(MSR_AMD64_TSC_RATIO, TSC_RATIO_DEFAULT);

755
	cpu_svm_disable();
756 757

	amd_pmu_disable_virt();
A
Avi Kivity 已提交
758 759
}

760
static int svm_hardware_enable(void)
A
Avi Kivity 已提交
761 762
{

763
	struct svm_cpu_data *sd;
A
Avi Kivity 已提交
764 765 766 767
	uint64_t efer;
	struct desc_struct *gdt;
	int me = raw_smp_processor_id();

768 769 770 771
	rdmsrl(MSR_EFER, efer);
	if (efer & EFER_SVME)
		return -EBUSY;

A
Avi Kivity 已提交
772
	if (!has_svm()) {
773
		pr_err("%s: err EOPNOTSUPP on %d\n", __func__, me);
774
		return -EINVAL;
A
Avi Kivity 已提交
775
	}
776 777
	sd = per_cpu(svm_data, me);
	if (!sd) {
778
		pr_err("%s: svm_data is NULL on %d\n", __func__, me);
779
		return -EINVAL;
A
Avi Kivity 已提交
780 781
	}

782 783 784
	sd->asid_generation = 1;
	sd->max_asid = cpuid_ebx(SVM_CPUID_FUNC) - 1;
	sd->next_asid = sd->max_asid + 1;
A
Avi Kivity 已提交
785

786
	gdt = get_current_gdt_rw();
787
	sd->tss_desc = (struct kvm_ldttss_desc *)(gdt + GDT_ENTRY_TSS);
A
Avi Kivity 已提交
788

789
	wrmsrl(MSR_EFER, efer | EFER_SVME);
A
Avi Kivity 已提交
790

791
	wrmsrl(MSR_VM_HSAVE_PA, page_to_pfn(sd->save_area) << PAGE_SHIFT);
792

793 794
	if (static_cpu_has(X86_FEATURE_TSCRATEMSR)) {
		wrmsrl(MSR_AMD64_TSC_RATIO, TSC_RATIO_DEFAULT);
795
		__this_cpu_write(current_tsc_ratio, TSC_RATIO_DEFAULT);
796 797
	}

798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827

	/*
	 * Get OSVW bits.
	 *
	 * Note that it is possible to have a system with mixed processor
	 * revisions and therefore different OSVW bits. If bits are not the same
	 * on different processors then choose the worst case (i.e. if erratum
	 * is present on one processor and not on another then assume that the
	 * erratum is present everywhere).
	 */
	if (cpu_has(&boot_cpu_data, X86_FEATURE_OSVW)) {
		uint64_t len, status = 0;
		int err;

		len = native_read_msr_safe(MSR_AMD64_OSVW_ID_LENGTH, &err);
		if (!err)
			status = native_read_msr_safe(MSR_AMD64_OSVW_STATUS,
						      &err);

		if (err)
			osvw_status = osvw_len = 0;
		else {
			if (len < osvw_len)
				osvw_len = len;
			osvw_status |= status;
			osvw_status &= (1ULL << osvw_len) - 1;
		}
	} else
		osvw_status = osvw_len = 0;

828 829
	svm_init_erratum_383();

830 831
	amd_pmu_enable_virt();

832
	return 0;
A
Avi Kivity 已提交
833 834
}

835 836
static void svm_cpu_uninit(int cpu)
{
837
	struct svm_cpu_data *sd = per_cpu(svm_data, raw_smp_processor_id());
838

839
	if (!sd)
840 841 842
		return;

	per_cpu(svm_data, raw_smp_processor_id()) = NULL;
843 844
	__free_page(sd->save_area);
	kfree(sd);
845 846
}

A
Avi Kivity 已提交
847 848
static int svm_cpu_init(int cpu)
{
849
	struct svm_cpu_data *sd;
A
Avi Kivity 已提交
850 851
	int r;

852 853
	sd = kzalloc(sizeof(struct svm_cpu_data), GFP_KERNEL);
	if (!sd)
A
Avi Kivity 已提交
854
		return -ENOMEM;
855 856
	sd->cpu = cpu;
	sd->save_area = alloc_page(GFP_KERNEL);
A
Avi Kivity 已提交
857
	r = -ENOMEM;
858
	if (!sd->save_area)
A
Avi Kivity 已提交
859 860
		goto err_1;

861
	per_cpu(svm_data, cpu) = sd;
A
Avi Kivity 已提交
862 863 864 865

	return 0;

err_1:
866
	kfree(sd);
A
Avi Kivity 已提交
867 868 869 870
	return r;

}

871 872 873 874 875 876 877 878 879 880 881
static bool valid_msr_intercept(u32 index)
{
	int i;

	for (i = 0; direct_access_msrs[i].index != MSR_INVALID; i++)
		if (direct_access_msrs[i].index == index)
			return true;

	return false;
}

882 883
static void set_msr_interception(u32 *msrpm, unsigned msr,
				 int read, int write)
A
Avi Kivity 已提交
884
{
885 886 887
	u8 bit_read, bit_write;
	unsigned long tmp;
	u32 offset;
A
Avi Kivity 已提交
888

889 890 891 892 893 894
	/*
	 * If this warning triggers extend the direct_access_msrs list at the
	 * beginning of the file
	 */
	WARN_ON(!valid_msr_intercept(msr));

895 896 897 898 899 900 901 902 903 904 905
	offset    = svm_msrpm_offset(msr);
	bit_read  = 2 * (msr & 0x0f);
	bit_write = 2 * (msr & 0x0f) + 1;
	tmp       = msrpm[offset];

	BUG_ON(offset == MSR_INVALID);

	read  ? clear_bit(bit_read,  &tmp) : set_bit(bit_read,  &tmp);
	write ? clear_bit(bit_write, &tmp) : set_bit(bit_write, &tmp);

	msrpm[offset] = tmp;
A
Avi Kivity 已提交
906 907
}

908
static void svm_vcpu_init_msrpm(u32 *msrpm)
A
Avi Kivity 已提交
909 910 911
{
	int i;

912 913
	memset(msrpm, 0xff, PAGE_SIZE * (1 << MSRPM_ALLOC_ORDER));

914 915 916 917 918 919
	for (i = 0; direct_access_msrs[i].index != MSR_INVALID; i++) {
		if (!direct_access_msrs[i].always)
			continue;

		set_msr_interception(msrpm, direct_access_msrs[i].index, 1, 1);
	}
920 921
}

922 923 924 925 926 927 928 929
static void add_msr_offset(u32 offset)
{
	int i;

	for (i = 0; i < MSRPM_OFFSETS; ++i) {

		/* Offset already in list? */
		if (msrpm_offsets[i] == offset)
930
			return;
931 932 933 934 935 936 937 938 939

		/* Slot used by another offset? */
		if (msrpm_offsets[i] != MSR_INVALID)
			continue;

		/* Add offset to list */
		msrpm_offsets[i] = offset;

		return;
A
Avi Kivity 已提交
940
	}
941 942 943 944 945

	/*
	 * If this BUG triggers the msrpm_offsets table has an overflow. Just
	 * increase MSRPM_OFFSETS in this case.
	 */
946
	BUG();
A
Avi Kivity 已提交
947 948
}

949
static void init_msrpm_offsets(void)
950
{
951
	int i;
952

953 954 955 956 957 958 959 960 961 962
	memset(msrpm_offsets, 0xff, sizeof(msrpm_offsets));

	for (i = 0; direct_access_msrs[i].index != MSR_INVALID; i++) {
		u32 offset;

		offset = svm_msrpm_offset(direct_access_msrs[i].index);
		BUG_ON(offset == MSR_INVALID);

		add_msr_offset(offset);
	}
963 964
}

965 966 967 968
static void svm_enable_lbrv(struct vcpu_svm *svm)
{
	u32 *msrpm = svm->msrpm;

969
	svm->vmcb->control.virt_ext |= LBR_CTL_ENABLE_MASK;
970 971 972 973 974 975 976 977 978 979
	set_msr_interception(msrpm, MSR_IA32_LASTBRANCHFROMIP, 1, 1);
	set_msr_interception(msrpm, MSR_IA32_LASTBRANCHTOIP, 1, 1);
	set_msr_interception(msrpm, MSR_IA32_LASTINTFROMIP, 1, 1);
	set_msr_interception(msrpm, MSR_IA32_LASTINTTOIP, 1, 1);
}

static void svm_disable_lbrv(struct vcpu_svm *svm)
{
	u32 *msrpm = svm->msrpm;

980
	svm->vmcb->control.virt_ext &= ~LBR_CTL_ENABLE_MASK;
981 982 983 984 985 986
	set_msr_interception(msrpm, MSR_IA32_LASTBRANCHFROMIP, 0, 0);
	set_msr_interception(msrpm, MSR_IA32_LASTBRANCHTOIP, 0, 0);
	set_msr_interception(msrpm, MSR_IA32_LASTINTFROMIP, 0, 0);
	set_msr_interception(msrpm, MSR_IA32_LASTINTTOIP, 0, 0);
}

987 988 989
static void disable_nmi_singlestep(struct vcpu_svm *svm)
{
	svm->nmi_singlestep = false;
990

991 992 993 994 995 996 997
	if (!(svm->vcpu.guest_debug & KVM_GUESTDBG_SINGLESTEP)) {
		/* Clear our flags if they were not set by the guest */
		if (!(svm->nmi_singlestep_guest_rflags & X86_EFLAGS_TF))
			svm->vmcb->save.rflags &= ~X86_EFLAGS_TF;
		if (!(svm->nmi_singlestep_guest_rflags & X86_EFLAGS_RF))
			svm->vmcb->save.rflags &= ~X86_EFLAGS_RF;
	}
998 999
}

1000 1001 1002 1003 1004 1005
/* Note:
 * This hash table is used to map VM_ID to a struct kvm_arch,
 * when handling AMD IOMMU GALOG notification to schedule in
 * a particular vCPU.
 */
#define SVM_VM_DATA_HASH_BITS	8
1006
static DEFINE_HASHTABLE(svm_vm_data_hash, SVM_VM_DATA_HASH_BITS);
1007 1008
static u32 next_vm_id = 0;
static bool next_vm_id_wrapped = 0;
1009
static DEFINE_SPINLOCK(svm_vm_data_hash_lock);
1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050

/* Note:
 * This function is called from IOMMU driver to notify
 * SVM to schedule in a particular vCPU of a particular VM.
 */
static int avic_ga_log_notifier(u32 ga_tag)
{
	unsigned long flags;
	struct kvm_arch *ka = NULL;
	struct kvm_vcpu *vcpu = NULL;
	u32 vm_id = AVIC_GATAG_TO_VMID(ga_tag);
	u32 vcpu_id = AVIC_GATAG_TO_VCPUID(ga_tag);

	pr_debug("SVM: %s: vm_id=%#x, vcpu_id=%#x\n", __func__, vm_id, vcpu_id);

	spin_lock_irqsave(&svm_vm_data_hash_lock, flags);
	hash_for_each_possible(svm_vm_data_hash, ka, hnode, vm_id) {
		struct kvm *kvm = container_of(ka, struct kvm, arch);
		struct kvm_arch *vm_data = &kvm->arch;

		if (vm_data->avic_vm_id != vm_id)
			continue;
		vcpu = kvm_get_vcpu_by_id(kvm, vcpu_id);
		break;
	}
	spin_unlock_irqrestore(&svm_vm_data_hash_lock, flags);

	if (!vcpu)
		return 0;

	/* Note:
	 * At this point, the IOMMU should have already set the pending
	 * bit in the vAPIC backing page. So, we just need to schedule
	 * in the vcpu.
	 */
	if (vcpu->mode == OUTSIDE_GUEST_MODE)
		kvm_vcpu_wake_up(vcpu);

	return 0;
}

A
Avi Kivity 已提交
1051 1052 1053 1054
static __init int svm_hardware_setup(void)
{
	int cpu;
	struct page *iopm_pages;
1055
	void *iopm_va;
A
Avi Kivity 已提交
1056 1057 1058 1059 1060 1061
	int r;

	iopm_pages = alloc_pages(GFP_KERNEL, IOPM_ALLOC_ORDER);

	if (!iopm_pages)
		return -ENOMEM;
1062 1063 1064

	iopm_va = page_address(iopm_pages);
	memset(iopm_va, 0xff, PAGE_SIZE * (1 << IOPM_ALLOC_ORDER));
A
Avi Kivity 已提交
1065 1066
	iopm_base = page_to_pfn(iopm_pages) << PAGE_SHIFT;

1067 1068
	init_msrpm_offsets();

1069 1070 1071
	if (boot_cpu_has(X86_FEATURE_NX))
		kvm_enable_efer_bits(EFER_NX);

A
Alexander Graf 已提交
1072 1073 1074
	if (boot_cpu_has(X86_FEATURE_FXSR_OPT))
		kvm_enable_efer_bits(EFER_FFXSR);

1075 1076
	if (boot_cpu_has(X86_FEATURE_TSCRATEMSR)) {
		kvm_has_tsc_control = true;
1077 1078
		kvm_max_tsc_scaling_ratio = TSC_RATIO_MAX;
		kvm_tsc_scaling_ratio_frac_bits = 32;
1079 1080
	}

1081 1082
	if (nested) {
		printk(KERN_INFO "kvm: Nested Virtualization enabled\n");
1083
		kvm_enable_efer_bits(EFER_SVME | EFER_LMSLE);
1084 1085
	}

Z
Zachary Amsden 已提交
1086
	for_each_possible_cpu(cpu) {
A
Avi Kivity 已提交
1087 1088
		r = svm_cpu_init(cpu);
		if (r)
1089
			goto err;
A
Avi Kivity 已提交
1090
	}
1091

1092
	if (!boot_cpu_has(X86_FEATURE_NPT))
1093 1094
		npt_enabled = false;

1095 1096 1097 1098 1099
	if (npt_enabled && !npt) {
		printk(KERN_INFO "kvm: Nested Paging disabled\n");
		npt_enabled = false;
	}

1100
	if (npt_enabled) {
1101
		printk(KERN_INFO "kvm: Nested Paging enabled\n");
1102
		kvm_enable_tdp();
1103 1104
	} else
		kvm_disable_tdp();
1105

1106 1107 1108
	if (avic) {
		if (!npt_enabled ||
		    !boot_cpu_has(X86_FEATURE_AVIC) ||
1109
		    !IS_ENABLED(CONFIG_X86_LOCAL_APIC)) {
1110
			avic = false;
1111
		} else {
1112
			pr_info("AVIC enabled\n");
1113 1114 1115

			amd_iommu_register_ga_log_notifier(&avic_ga_log_notifier);
		}
1116
	}
1117

1118 1119
	if (vls) {
		if (!npt_enabled ||
1120
		    !boot_cpu_has(X86_FEATURE_V_VMSAVE_VMLOAD) ||
1121 1122 1123 1124 1125 1126 1127
		    !IS_ENABLED(CONFIG_X86_64)) {
			vls = false;
		} else {
			pr_info("Virtual VMLOAD VMSAVE supported\n");
		}
	}

1128 1129 1130 1131 1132 1133 1134
	if (vgif) {
		if (!boot_cpu_has(X86_FEATURE_VGIF))
			vgif = false;
		else
			pr_info("Virtual GIF supported\n");
	}

A
Avi Kivity 已提交
1135 1136
	return 0;

1137
err:
A
Avi Kivity 已提交
1138 1139 1140 1141 1142 1143 1144
	__free_pages(iopm_pages, IOPM_ALLOC_ORDER);
	iopm_base = 0;
	return r;
}

static __exit void svm_hardware_unsetup(void)
{
1145 1146
	int cpu;

Z
Zachary Amsden 已提交
1147
	for_each_possible_cpu(cpu)
1148 1149
		svm_cpu_uninit(cpu);

A
Avi Kivity 已提交
1150
	__free_pages(pfn_to_page(iopm_base >> PAGE_SHIFT), IOPM_ALLOC_ORDER);
1151
	iopm_base = 0;
A
Avi Kivity 已提交
1152 1153 1154 1155 1156 1157
}

static void init_seg(struct vmcb_seg *seg)
{
	seg->selector = 0;
	seg->attrib = SVM_SELECTOR_P_MASK | SVM_SELECTOR_S_MASK |
J
Joerg Roedel 已提交
1158
		      SVM_SELECTOR_WRITE_MASK; /* Read/Write Data Segment */
A
Avi Kivity 已提交
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170
	seg->limit = 0xffff;
	seg->base = 0;
}

static void init_sys_seg(struct vmcb_seg *seg, uint32_t type)
{
	seg->selector = 0;
	seg->attrib = SVM_SELECTOR_P_MASK | type;
	seg->limit = 0xffff;
	seg->base = 0;
}

1171 1172 1173 1174 1175
static void svm_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u64 g_tsc_offset = 0;

1176
	if (is_guest_mode(vcpu)) {
1177 1178 1179
		g_tsc_offset = svm->vmcb->control.tsc_offset -
			       svm->nested.hsave->control.tsc_offset;
		svm->nested.hsave->control.tsc_offset = offset;
1180 1181 1182 1183
	} else
		trace_kvm_write_tsc_offset(vcpu->vcpu_id,
					   svm->vmcb->control.tsc_offset,
					   offset);
1184 1185

	svm->vmcb->control.tsc_offset = offset + g_tsc_offset;
1186 1187

	mark_dirty(svm->vmcb, VMCB_INTERCEPTS);
1188 1189
}

1190 1191 1192 1193
static void avic_init_vmcb(struct vcpu_svm *svm)
{
	struct vmcb *vmcb = svm->vmcb;
	struct kvm_arch *vm_data = &svm->vcpu.kvm->arch;
1194 1195 1196
	phys_addr_t bpa = __sme_set(page_to_phys(svm->avic_backing_page));
	phys_addr_t lpa = __sme_set(page_to_phys(vm_data->avic_logical_id_table_page));
	phys_addr_t ppa = __sme_set(page_to_phys(vm_data->avic_physical_id_table_page));
1197 1198 1199 1200 1201 1202 1203 1204 1205

	vmcb->control.avic_backing_page = bpa & AVIC_HPA_MASK;
	vmcb->control.avic_logical_id = lpa & AVIC_HPA_MASK;
	vmcb->control.avic_physical_id = ppa & AVIC_HPA_MASK;
	vmcb->control.avic_physical_id |= AVIC_MAX_PHYSICAL_ID_COUNT;
	vmcb->control.int_ctl |= AVIC_ENABLE_MASK;
	svm->vcpu.arch.apicv_active = true;
}

P
Paolo Bonzini 已提交
1206
static void init_vmcb(struct vcpu_svm *svm)
A
Avi Kivity 已提交
1207
{
1208 1209
	struct vmcb_control_area *control = &svm->vmcb->control;
	struct vmcb_save_area *save = &svm->vmcb->save;
A
Avi Kivity 已提交
1210

1211
	svm->vcpu.arch.hflags = 0;
1212

1213 1214 1215 1216 1217 1218
	set_cr_intercept(svm, INTERCEPT_CR0_READ);
	set_cr_intercept(svm, INTERCEPT_CR3_READ);
	set_cr_intercept(svm, INTERCEPT_CR4_READ);
	set_cr_intercept(svm, INTERCEPT_CR0_WRITE);
	set_cr_intercept(svm, INTERCEPT_CR3_WRITE);
	set_cr_intercept(svm, INTERCEPT_CR4_WRITE);
1219 1220
	if (!kvm_vcpu_apicv_active(&svm->vcpu))
		set_cr_intercept(svm, INTERCEPT_CR8_WRITE);
A
Avi Kivity 已提交
1221

1222
	set_dr_intercepts(svm);
A
Avi Kivity 已提交
1223

1224 1225 1226
	set_exception_intercept(svm, PF_VECTOR);
	set_exception_intercept(svm, UD_VECTOR);
	set_exception_intercept(svm, MC_VECTOR);
1227
	set_exception_intercept(svm, AC_VECTOR);
1228
	set_exception_intercept(svm, DB_VECTOR);
A
Avi Kivity 已提交
1229

1230 1231 1232 1233
	set_intercept(svm, INTERCEPT_INTR);
	set_intercept(svm, INTERCEPT_NMI);
	set_intercept(svm, INTERCEPT_SMI);
	set_intercept(svm, INTERCEPT_SELECTIVE_CR0);
A
Avi Kivity 已提交
1234
	set_intercept(svm, INTERCEPT_RDPMC);
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
	set_intercept(svm, INTERCEPT_CPUID);
	set_intercept(svm, INTERCEPT_INVD);
	set_intercept(svm, INTERCEPT_HLT);
	set_intercept(svm, INTERCEPT_INVLPG);
	set_intercept(svm, INTERCEPT_INVLPGA);
	set_intercept(svm, INTERCEPT_IOIO_PROT);
	set_intercept(svm, INTERCEPT_MSR_PROT);
	set_intercept(svm, INTERCEPT_TASK_SWITCH);
	set_intercept(svm, INTERCEPT_SHUTDOWN);
	set_intercept(svm, INTERCEPT_VMRUN);
	set_intercept(svm, INTERCEPT_VMMCALL);
	set_intercept(svm, INTERCEPT_VMLOAD);
	set_intercept(svm, INTERCEPT_VMSAVE);
	set_intercept(svm, INTERCEPT_STGI);
	set_intercept(svm, INTERCEPT_CLGI);
	set_intercept(svm, INTERCEPT_SKINIT);
	set_intercept(svm, INTERCEPT_WBINVD);
J
Joerg Roedel 已提交
1252
	set_intercept(svm, INTERCEPT_XSETBV);
A
Avi Kivity 已提交
1253

1254 1255 1256 1257 1258
	if (!kvm_mwait_in_guest()) {
		set_intercept(svm, INTERCEPT_MONITOR);
		set_intercept(svm, INTERCEPT_MWAIT);
	}

1259 1260
	control->iopm_base_pa = __sme_set(iopm_base);
	control->msrpm_base_pa = __sme_set(__pa(svm->msrpm));
A
Avi Kivity 已提交
1261 1262 1263 1264 1265 1266 1267 1268 1269
	control->int_ctl = V_INTR_MASKING_MASK;

	init_seg(&save->es);
	init_seg(&save->ss);
	init_seg(&save->ds);
	init_seg(&save->fs);
	init_seg(&save->gs);

	save->cs.selector = 0xf000;
1270
	save->cs.base = 0xffff0000;
A
Avi Kivity 已提交
1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281
	/* Executable/Readable Code Segment */
	save->cs.attrib = SVM_SELECTOR_READ_MASK | SVM_SELECTOR_P_MASK |
		SVM_SELECTOR_S_MASK | SVM_SELECTOR_CODE_MASK;
	save->cs.limit = 0xffff;

	save->gdtr.limit = 0xffff;
	save->idtr.limit = 0xffff;

	init_sys_seg(&save->ldtr, SEG_TYPE_LDT);
	init_sys_seg(&save->tr, SEG_TYPE_BUSY_TSS16);

P
Paolo Bonzini 已提交
1282
	svm_set_efer(&svm->vcpu, 0);
M
Mike Day 已提交
1283
	save->dr6 = 0xffff0ff0;
1284
	kvm_set_rflags(&svm->vcpu, 2);
A
Avi Kivity 已提交
1285
	save->rip = 0x0000fff0;
1286
	svm->vcpu.arch.regs[VCPU_REGS_RIP] = save->rip;
A
Avi Kivity 已提交
1287

J
Joerg Roedel 已提交
1288
	/*
1289
	 * svm_set_cr0() sets PG and WP and clears NW and CD on save->cr0.
1290
	 * It also updates the guest-visible cr0 value.
A
Avi Kivity 已提交
1291
	 */
1292
	svm_set_cr0(&svm->vcpu, X86_CR0_NW | X86_CR0_CD | X86_CR0_ET);
1293
	kvm_mmu_reset_context(&svm->vcpu);
1294

1295
	save->cr4 = X86_CR4_PAE;
A
Avi Kivity 已提交
1296
	/* rdx = ?? */
1297 1298 1299 1300

	if (npt_enabled) {
		/* Setup VMCB for Nested Paging */
		control->nested_ctl = 1;
1301
		clr_intercept(svm, INTERCEPT_INVLPG);
1302
		clr_exception_intercept(svm, PF_VECTOR);
1303 1304
		clr_cr_intercept(svm, INTERCEPT_CR3_READ);
		clr_cr_intercept(svm, INTERCEPT_CR3_WRITE);
1305
		save->g_pat = svm->vcpu.arch.pat;
1306 1307 1308
		save->cr3 = 0;
		save->cr4 = 0;
	}
1309
	svm->asid_generation = 0;
1310

1311
	svm->nested.vmcb = 0;
1312 1313
	svm->vcpu.arch.hflags = 0;

1314
	if (boot_cpu_has(X86_FEATURE_PAUSEFILTER)) {
1315
		control->pause_filter_count = 3000;
1316
		set_intercept(svm, INTERCEPT_PAUSE);
1317 1318
	}

1319 1320 1321
	if (avic)
		avic_init_vmcb(svm);

1322 1323 1324 1325 1326 1327 1328 1329 1330 1331
	/*
	 * If hardware supports Virtual VMLOAD VMSAVE then enable it
	 * in VMCB and clear intercepts to avoid #VMEXIT.
	 */
	if (vls) {
		clr_intercept(svm, INTERCEPT_VMLOAD);
		clr_intercept(svm, INTERCEPT_VMSAVE);
		svm->vmcb->control.virt_ext |= VIRTUAL_VMLOAD_VMSAVE_ENABLE_MASK;
	}

1332 1333 1334 1335 1336 1337
	if (vgif) {
		clr_intercept(svm, INTERCEPT_STGI);
		clr_intercept(svm, INTERCEPT_CLGI);
		svm->vmcb->control.int_ctl |= V_GIF_ENABLE_MASK;
	}

1338 1339
	mark_all_dirty(svm->vmcb);

1340
	enable_gif(svm);
1341 1342 1343

}

1344 1345
static u64 *avic_get_physical_id_entry(struct kvm_vcpu *vcpu,
				       unsigned int index)
1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409
{
	u64 *avic_physical_id_table;
	struct kvm_arch *vm_data = &vcpu->kvm->arch;

	if (index >= AVIC_MAX_PHYSICAL_ID_COUNT)
		return NULL;

	avic_physical_id_table = page_address(vm_data->avic_physical_id_table_page);

	return &avic_physical_id_table[index];
}

/**
 * Note:
 * AVIC hardware walks the nested page table to check permissions,
 * but does not use the SPA address specified in the leaf page
 * table entry since it uses  address in the AVIC_BACKING_PAGE pointer
 * field of the VMCB. Therefore, we set up the
 * APIC_ACCESS_PAGE_PRIVATE_MEMSLOT (4KB) here.
 */
static int avic_init_access_page(struct kvm_vcpu *vcpu)
{
	struct kvm *kvm = vcpu->kvm;
	int ret;

	if (kvm->arch.apic_access_page_done)
		return 0;

	ret = x86_set_memory_region(kvm,
				    APIC_ACCESS_PAGE_PRIVATE_MEMSLOT,
				    APIC_DEFAULT_PHYS_BASE,
				    PAGE_SIZE);
	if (ret)
		return ret;

	kvm->arch.apic_access_page_done = true;
	return 0;
}

static int avic_init_backing_page(struct kvm_vcpu *vcpu)
{
	int ret;
	u64 *entry, new_entry;
	int id = vcpu->vcpu_id;
	struct vcpu_svm *svm = to_svm(vcpu);

	ret = avic_init_access_page(vcpu);
	if (ret)
		return ret;

	if (id >= AVIC_MAX_PHYSICAL_ID_COUNT)
		return -EINVAL;

	if (!svm->vcpu.arch.apic->regs)
		return -EINVAL;

	svm->avic_backing_page = virt_to_page(svm->vcpu.arch.apic->regs);

	/* Setting AVIC backing page address in the phy APIC ID table */
	entry = avic_get_physical_id_entry(vcpu, id);
	if (!entry)
		return -EINVAL;

	new_entry = READ_ONCE(*entry);
1410 1411 1412
	new_entry = __sme_set((page_to_phys(svm->avic_backing_page) &
			      AVIC_PHYSICAL_ID_ENTRY_BACKING_PAGE_MASK) |
			      AVIC_PHYSICAL_ID_ENTRY_VALID_MASK);
1413 1414 1415 1416 1417 1418 1419 1420 1421
	WRITE_ONCE(*entry, new_entry);

	svm->avic_physical_id_cache = entry;

	return 0;
}

static void avic_vm_destroy(struct kvm *kvm)
{
1422
	unsigned long flags;
1423 1424
	struct kvm_arch *vm_data = &kvm->arch;

1425 1426 1427
	if (!avic)
		return;

1428 1429 1430 1431
	if (vm_data->avic_logical_id_table_page)
		__free_page(vm_data->avic_logical_id_table_page);
	if (vm_data->avic_physical_id_table_page)
		__free_page(vm_data->avic_physical_id_table_page);
1432 1433 1434 1435

	spin_lock_irqsave(&svm_vm_data_hash_lock, flags);
	hash_del(&vm_data->hnode);
	spin_unlock_irqrestore(&svm_vm_data_hash_lock, flags);
1436 1437 1438 1439
}

static int avic_vm_init(struct kvm *kvm)
{
1440
	unsigned long flags;
1441
	int err = -ENOMEM;
1442 1443 1444
	struct kvm_arch *vm_data = &kvm->arch;
	struct page *p_page;
	struct page *l_page;
1445 1446
	struct kvm_arch *ka;
	u32 vm_id;
1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466

	if (!avic)
		return 0;

	/* Allocating physical APIC ID table (4KB) */
	p_page = alloc_page(GFP_KERNEL);
	if (!p_page)
		goto free_avic;

	vm_data->avic_physical_id_table_page = p_page;
	clear_page(page_address(p_page));

	/* Allocating logical APIC ID table (4KB) */
	l_page = alloc_page(GFP_KERNEL);
	if (!l_page)
		goto free_avic;

	vm_data->avic_logical_id_table_page = l_page;
	clear_page(page_address(l_page));

1467
	spin_lock_irqsave(&svm_vm_data_hash_lock, flags);
1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483
 again:
	vm_id = next_vm_id = (next_vm_id + 1) & AVIC_VM_ID_MASK;
	if (vm_id == 0) { /* id is 1-based, zero is not okay */
		next_vm_id_wrapped = 1;
		goto again;
	}
	/* Is it still in use? Only possible if wrapped at least once */
	if (next_vm_id_wrapped) {
		hash_for_each_possible(svm_vm_data_hash, ka, hnode, vm_id) {
			struct kvm *k2 = container_of(ka, struct kvm, arch);
			struct kvm_arch *vd2 = &k2->arch;
			if (vd2->avic_vm_id == vm_id)
				goto again;
		}
	}
	vm_data->avic_vm_id = vm_id;
1484 1485 1486
	hash_add(svm_vm_data_hash, &vm_data->hnode, vm_data->avic_vm_id);
	spin_unlock_irqrestore(&svm_vm_data_hash_lock, flags);

1487 1488 1489 1490 1491
	return 0;

free_avic:
	avic_vm_destroy(kvm);
	return err;
A
Avi Kivity 已提交
1492 1493
}

1494 1495
static inline int
avic_update_iommu_vcpu_affinity(struct kvm_vcpu *vcpu, int cpu, bool r)
1496
{
1497 1498 1499
	int ret = 0;
	unsigned long flags;
	struct amd_svm_iommu_ir *ir;
1500 1501
	struct vcpu_svm *svm = to_svm(vcpu);

1502 1503
	if (!kvm_arch_has_assigned_device(vcpu->kvm))
		return 0;
1504

1505 1506 1507 1508 1509
	/*
	 * Here, we go through the per-vcpu ir_list to update all existing
	 * interrupt remapping table entry targeting this vcpu.
	 */
	spin_lock_irqsave(&svm->ir_list_lock, flags);
1510

1511 1512
	if (list_empty(&svm->ir_list))
		goto out;
1513

1514 1515 1516 1517 1518 1519 1520 1521
	list_for_each_entry(ir, &svm->ir_list, node) {
		ret = amd_iommu_update_ga(cpu, r, ir->data);
		if (ret)
			break;
	}
out:
	spin_unlock_irqrestore(&svm->ir_list_lock, flags);
	return ret;
1522 1523 1524 1525 1526 1527
}

static void avic_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
	u64 entry;
	/* ID = 0xff (broadcast), ID > 0xff (reserved) */
1528
	int h_physical_id = kvm_cpu_get_apicid(cpu);
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547
	struct vcpu_svm *svm = to_svm(vcpu);

	if (!kvm_vcpu_apicv_active(vcpu))
		return;

	if (WARN_ON(h_physical_id >= AVIC_MAX_PHYSICAL_ID_COUNT))
		return;

	entry = READ_ONCE(*(svm->avic_physical_id_cache));
	WARN_ON(entry & AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK);

	entry &= ~AVIC_PHYSICAL_ID_ENTRY_HOST_PHYSICAL_ID_MASK;
	entry |= (h_physical_id & AVIC_PHYSICAL_ID_ENTRY_HOST_PHYSICAL_ID_MASK);

	entry &= ~AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK;
	if (svm->avic_is_running)
		entry |= AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK;

	WRITE_ONCE(*(svm->avic_physical_id_cache), entry);
1548 1549
	avic_update_iommu_vcpu_affinity(vcpu, h_physical_id,
					svm->avic_is_running);
1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560
}

static void avic_vcpu_put(struct kvm_vcpu *vcpu)
{
	u64 entry;
	struct vcpu_svm *svm = to_svm(vcpu);

	if (!kvm_vcpu_apicv_active(vcpu))
		return;

	entry = READ_ONCE(*(svm->avic_physical_id_cache));
1561 1562 1563
	if (entry & AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK)
		avic_update_iommu_vcpu_affinity(vcpu, -1, 0);

1564 1565
	entry &= ~AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK;
	WRITE_ONCE(*(svm->avic_physical_id_cache), entry);
A
Avi Kivity 已提交
1566 1567
}

1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581
/**
 * This function is called during VCPU halt/unhalt.
 */
static void avic_set_running(struct kvm_vcpu *vcpu, bool is_run)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	svm->avic_is_running = is_run;
	if (is_run)
		avic_vcpu_load(vcpu, vcpu->cpu);
	else
		avic_vcpu_put(vcpu);
}

1582
static void svm_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
1583 1584
{
	struct vcpu_svm *svm = to_svm(vcpu);
1585 1586
	u32 dummy;
	u32 eax = 1;
1587

1588 1589 1590 1591 1592 1593
	if (!init_event) {
		svm->vcpu.arch.apic_base = APIC_DEFAULT_PHYS_BASE |
					   MSR_IA32_APICBASE_ENABLE;
		if (kvm_vcpu_is_reset_bsp(&svm->vcpu))
			svm->vcpu.arch.apic_base |= MSR_IA32_APICBASE_BSP;
	}
P
Paolo Bonzini 已提交
1594
	init_vmcb(svm);
A
Avi Kivity 已提交
1595

1596
	kvm_cpuid(vcpu, &eax, &dummy, &dummy, &dummy, true);
1597
	kvm_register_write(vcpu, VCPU_REGS_RDX, eax);
1598 1599 1600

	if (kvm_vcpu_apicv_active(vcpu) && !init_event)
		avic_update_vapic_bar(svm, APIC_DEFAULT_PHYS_BASE);
1601 1602
}

R
Rusty Russell 已提交
1603
static struct kvm_vcpu *svm_create_vcpu(struct kvm *kvm, unsigned int id)
A
Avi Kivity 已提交
1604
{
1605
	struct vcpu_svm *svm;
A
Avi Kivity 已提交
1606
	struct page *page;
1607
	struct page *msrpm_pages;
A
Alexander Graf 已提交
1608
	struct page *hsave_page;
A
Alexander Graf 已提交
1609
	struct page *nested_msrpm_pages;
R
Rusty Russell 已提交
1610
	int err;
A
Avi Kivity 已提交
1611

1612
	svm = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL);
R
Rusty Russell 已提交
1613 1614 1615 1616 1617 1618 1619 1620 1621
	if (!svm) {
		err = -ENOMEM;
		goto out;
	}

	err = kvm_vcpu_init(&svm->vcpu, kvm, id);
	if (err)
		goto free_svm;

1622
	err = -ENOMEM;
A
Avi Kivity 已提交
1623
	page = alloc_page(GFP_KERNEL);
1624
	if (!page)
R
Rusty Russell 已提交
1625
		goto uninit;
A
Avi Kivity 已提交
1626

1627 1628
	msrpm_pages = alloc_pages(GFP_KERNEL, MSRPM_ALLOC_ORDER);
	if (!msrpm_pages)
1629
		goto free_page1;
A
Alexander Graf 已提交
1630 1631 1632

	nested_msrpm_pages = alloc_pages(GFP_KERNEL, MSRPM_ALLOC_ORDER);
	if (!nested_msrpm_pages)
1633
		goto free_page2;
1634

A
Alexander Graf 已提交
1635 1636
	hsave_page = alloc_page(GFP_KERNEL);
	if (!hsave_page)
1637 1638
		goto free_page3;

1639 1640 1641 1642
	if (avic) {
		err = avic_init_backing_page(&svm->vcpu);
		if (err)
			goto free_page4;
1643 1644 1645

		INIT_LIST_HEAD(&svm->ir_list);
		spin_lock_init(&svm->ir_list_lock);
1646 1647
	}

1648 1649 1650 1651 1652
	/* We initialize this flag to true to make sure that the is_running
	 * bit would be set the first time the vcpu is loaded.
	 */
	svm->avic_is_running = true;

1653
	svm->nested.hsave = page_address(hsave_page);
A
Alexander Graf 已提交
1654

1655 1656 1657
	svm->msrpm = page_address(msrpm_pages);
	svm_vcpu_init_msrpm(svm->msrpm);

1658
	svm->nested.msrpm = page_address(nested_msrpm_pages);
1659
	svm_vcpu_init_msrpm(svm->nested.msrpm);
A
Alexander Graf 已提交
1660

1661 1662
	svm->vmcb = page_address(page);
	clear_page(svm->vmcb);
1663
	svm->vmcb_pa = __sme_set(page_to_pfn(page) << PAGE_SHIFT);
1664
	svm->asid_generation = 0;
P
Paolo Bonzini 已提交
1665
	init_vmcb(svm);
A
Avi Kivity 已提交
1666

1667 1668
	svm_init_osvw(&svm->vcpu);

R
Rusty Russell 已提交
1669
	return &svm->vcpu;
1670

1671 1672
free_page4:
	__free_page(hsave_page);
1673 1674 1675 1676 1677 1678
free_page3:
	__free_pages(nested_msrpm_pages, MSRPM_ALLOC_ORDER);
free_page2:
	__free_pages(msrpm_pages, MSRPM_ALLOC_ORDER);
free_page1:
	__free_page(page);
R
Rusty Russell 已提交
1679 1680 1681
uninit:
	kvm_vcpu_uninit(&svm->vcpu);
free_svm:
1682
	kmem_cache_free(kvm_vcpu_cache, svm);
R
Rusty Russell 已提交
1683 1684
out:
	return ERR_PTR(err);
A
Avi Kivity 已提交
1685 1686 1687 1688
}

static void svm_free_vcpu(struct kvm_vcpu *vcpu)
{
1689 1690
	struct vcpu_svm *svm = to_svm(vcpu);

1691
	__free_page(pfn_to_page(__sme_clr(svm->vmcb_pa) >> PAGE_SHIFT));
1692
	__free_pages(virt_to_page(svm->msrpm), MSRPM_ALLOC_ORDER);
1693 1694
	__free_page(virt_to_page(svm->nested.hsave));
	__free_pages(virt_to_page(svm->nested.msrpm), MSRPM_ALLOC_ORDER);
R
Rusty Russell 已提交
1695
	kvm_vcpu_uninit(vcpu);
1696
	kmem_cache_free(kvm_vcpu_cache, svm);
A
Avi Kivity 已提交
1697 1698
}

1699
static void svm_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
A
Avi Kivity 已提交
1700
{
1701
	struct vcpu_svm *svm = to_svm(vcpu);
1702
	int i;
1703 1704

	if (unlikely(cpu != vcpu->cpu)) {
1705
		svm->asid_generation = 0;
1706
		mark_all_dirty(svm->vmcb);
1707
	}
1708

1709 1710 1711
#ifdef CONFIG_X86_64
	rdmsrl(MSR_GS_BASE, to_svm(vcpu)->host.gs_base);
#endif
1712 1713 1714 1715
	savesegment(fs, svm->host.fs);
	savesegment(gs, svm->host.gs);
	svm->host.ldt = kvm_read_ldt();

1716
	for (i = 0; i < NR_HOST_SAVE_USER_MSRS; i++)
1717
		rdmsrl(host_save_user_msrs[i], svm->host_user_msrs[i]);
1718

1719 1720 1721 1722 1723 1724
	if (static_cpu_has(X86_FEATURE_TSCRATEMSR)) {
		u64 tsc_ratio = vcpu->arch.tsc_scaling_ratio;
		if (tsc_ratio != __this_cpu_read(current_tsc_ratio)) {
			__this_cpu_write(current_tsc_ratio, tsc_ratio);
			wrmsrl(MSR_AMD64_TSC_RATIO, tsc_ratio);
		}
1725
	}
P
Paolo Bonzini 已提交
1726 1727 1728
	/* This assumes that the kernel never uses MSR_TSC_AUX */
	if (static_cpu_has(X86_FEATURE_RDTSCP))
		wrmsrl(MSR_TSC_AUX, svm->tsc_aux);
1729 1730

	avic_vcpu_load(vcpu, cpu);
A
Avi Kivity 已提交
1731 1732 1733 1734
}

static void svm_vcpu_put(struct kvm_vcpu *vcpu)
{
1735
	struct vcpu_svm *svm = to_svm(vcpu);
1736 1737
	int i;

1738 1739
	avic_vcpu_put(vcpu);

1740
	++vcpu->stat.host_state_reload;
1741 1742 1743
	kvm_load_ldt(svm->host.ldt);
#ifdef CONFIG_X86_64
	loadsegment(fs, svm->host.fs);
1744
	wrmsrl(MSR_KERNEL_GS_BASE, current->thread.gsbase);
1745
	load_gs_index(svm->host.gs);
1746
#else
1747
#ifdef CONFIG_X86_32_LAZY_GS
1748
	loadsegment(gs, svm->host.gs);
1749
#endif
1750
#endif
1751
	for (i = 0; i < NR_HOST_SAVE_USER_MSRS; i++)
1752
		wrmsrl(host_save_user_msrs[i], svm->host_user_msrs[i]);
A
Avi Kivity 已提交
1753 1754
}

1755 1756 1757 1758 1759 1760 1761 1762 1763 1764
static void svm_vcpu_blocking(struct kvm_vcpu *vcpu)
{
	avic_set_running(vcpu, false);
}

static void svm_vcpu_unblocking(struct kvm_vcpu *vcpu)
{
	avic_set_running(vcpu, true);
}

A
Avi Kivity 已提交
1765 1766
static unsigned long svm_get_rflags(struct kvm_vcpu *vcpu)
{
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777
	struct vcpu_svm *svm = to_svm(vcpu);
	unsigned long rflags = svm->vmcb->save.rflags;

	if (svm->nmi_singlestep) {
		/* Hide our flags if they were not set by the guest */
		if (!(svm->nmi_singlestep_guest_rflags & X86_EFLAGS_TF))
			rflags &= ~X86_EFLAGS_TF;
		if (!(svm->nmi_singlestep_guest_rflags & X86_EFLAGS_RF))
			rflags &= ~X86_EFLAGS_RF;
	}
	return rflags;
A
Avi Kivity 已提交
1778 1779 1780 1781
}

static void svm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags)
{
1782 1783 1784
	if (to_svm(vcpu)->nmi_singlestep)
		rflags |= (X86_EFLAGS_TF | X86_EFLAGS_RF);

P
Paolo Bonzini 已提交
1785
       /*
A
Andrea Gelmini 已提交
1786
        * Any change of EFLAGS.VM is accompanied by a reload of SS
P
Paolo Bonzini 已提交
1787 1788 1789
        * (caused by either a task switch or an inter-privilege IRET),
        * so we do not need to update the CPL here.
        */
1790
	to_svm(vcpu)->vmcb->save.rflags = rflags;
A
Avi Kivity 已提交
1791 1792
}

A
Avi Kivity 已提交
1793 1794 1795 1796 1797
static void svm_cache_reg(struct kvm_vcpu *vcpu, enum kvm_reg reg)
{
	switch (reg) {
	case VCPU_EXREG_PDPTR:
		BUG_ON(!npt_enabled);
1798
		load_pdptrs(vcpu, vcpu->arch.walk_mmu, kvm_read_cr3(vcpu));
A
Avi Kivity 已提交
1799 1800 1801 1802 1803 1804
		break;
	default:
		BUG();
	}
}

1805 1806
static void svm_set_vintr(struct vcpu_svm *svm)
{
1807
	set_intercept(svm, INTERCEPT_VINTR);
1808 1809 1810 1811
}

static void svm_clear_vintr(struct vcpu_svm *svm)
{
1812
	clr_intercept(svm, INTERCEPT_VINTR);
1813 1814
}

A
Avi Kivity 已提交
1815 1816
static struct vmcb_seg *svm_seg(struct kvm_vcpu *vcpu, int seg)
{
1817
	struct vmcb_save_area *save = &to_svm(vcpu)->vmcb->save;
A
Avi Kivity 已提交
1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829

	switch (seg) {
	case VCPU_SREG_CS: return &save->cs;
	case VCPU_SREG_DS: return &save->ds;
	case VCPU_SREG_ES: return &save->es;
	case VCPU_SREG_FS: return &save->fs;
	case VCPU_SREG_GS: return &save->gs;
	case VCPU_SREG_SS: return &save->ss;
	case VCPU_SREG_TR: return &save->tr;
	case VCPU_SREG_LDTR: return &save->ldtr;
	}
	BUG();
A
Al Viro 已提交
1830
	return NULL;
A
Avi Kivity 已提交
1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854
}

static u64 svm_get_segment_base(struct kvm_vcpu *vcpu, int seg)
{
	struct vmcb_seg *s = svm_seg(vcpu, seg);

	return s->base;
}

static void svm_get_segment(struct kvm_vcpu *vcpu,
			    struct kvm_segment *var, int seg)
{
	struct vmcb_seg *s = svm_seg(vcpu, seg);

	var->base = s->base;
	var->limit = s->limit;
	var->selector = s->selector;
	var->type = s->attrib & SVM_SELECTOR_TYPE_MASK;
	var->s = (s->attrib >> SVM_SELECTOR_S_SHIFT) & 1;
	var->dpl = (s->attrib >> SVM_SELECTOR_DPL_SHIFT) & 3;
	var->present = (s->attrib >> SVM_SELECTOR_P_SHIFT) & 1;
	var->avl = (s->attrib >> SVM_SELECTOR_AVL_SHIFT) & 1;
	var->l = (s->attrib >> SVM_SELECTOR_L_SHIFT) & 1;
	var->db = (s->attrib >> SVM_SELECTOR_DB_SHIFT) & 1;
1855 1856 1857 1858 1859 1860 1861 1862 1863 1864

	/*
	 * AMD CPUs circa 2014 track the G bit for all segments except CS.
	 * However, the SVM spec states that the G bit is not observed by the
	 * CPU, and some VMware virtual CPUs drop the G bit for all segments.
	 * So let's synthesize a legal G bit for all segments, this helps
	 * running KVM nested. It also helps cross-vendor migration, because
	 * Intel's vmentry has a check on the 'G' bit.
	 */
	var->g = s->limit > 0xfffff;
1865

J
Joerg Roedel 已提交
1866 1867
	/*
	 * AMD's VMCB does not have an explicit unusable field, so emulate it
1868 1869
	 * for cross vendor migration purposes by "not present"
	 */
1870
	var->unusable = !var->present;
1871

1872 1873 1874 1875 1876 1877
	switch (seg) {
	case VCPU_SREG_TR:
		/*
		 * Work around a bug where the busy flag in the tr selector
		 * isn't exposed
		 */
1878
		var->type |= 0x2;
1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893
		break;
	case VCPU_SREG_DS:
	case VCPU_SREG_ES:
	case VCPU_SREG_FS:
	case VCPU_SREG_GS:
		/*
		 * The accessed bit must always be set in the segment
		 * descriptor cache, although it can be cleared in the
		 * descriptor, the cached bit always remains at 1. Since
		 * Intel has a check on this, set it here to support
		 * cross-vendor migration.
		 */
		if (!var->unusable)
			var->type |= 0x1;
		break;
1894
	case VCPU_SREG_SS:
J
Joerg Roedel 已提交
1895 1896
		/*
		 * On AMD CPUs sometimes the DB bit in the segment
1897 1898 1899 1900 1901 1902
		 * descriptor is left as 1, although the whole segment has
		 * been made unusable. Clear it here to pass an Intel VMX
		 * entry check when cross vendor migrating.
		 */
		if (var->unusable)
			var->db = 0;
1903
		/* This is symmetric with svm_set_segment() */
J
Jan Kiszka 已提交
1904
		var->dpl = to_svm(vcpu)->vmcb->save.cpl;
1905
		break;
1906
	}
A
Avi Kivity 已提交
1907 1908
}

1909 1910 1911 1912 1913 1914 1915
static int svm_get_cpl(struct kvm_vcpu *vcpu)
{
	struct vmcb_save_area *save = &to_svm(vcpu)->vmcb->save;

	return save->cpl;
}

1916
static void svm_get_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
1917
{
1918 1919
	struct vcpu_svm *svm = to_svm(vcpu);

1920 1921
	dt->size = svm->vmcb->save.idtr.limit;
	dt->address = svm->vmcb->save.idtr.base;
A
Avi Kivity 已提交
1922 1923
}

1924
static void svm_set_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
1925
{
1926 1927
	struct vcpu_svm *svm = to_svm(vcpu);

1928 1929
	svm->vmcb->save.idtr.limit = dt->size;
	svm->vmcb->save.idtr.base = dt->address ;
1930
	mark_dirty(svm->vmcb, VMCB_DT);
A
Avi Kivity 已提交
1931 1932
}

1933
static void svm_get_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
1934
{
1935 1936
	struct vcpu_svm *svm = to_svm(vcpu);

1937 1938
	dt->size = svm->vmcb->save.gdtr.limit;
	dt->address = svm->vmcb->save.gdtr.base;
A
Avi Kivity 已提交
1939 1940
}

1941
static void svm_set_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
1942
{
1943 1944
	struct vcpu_svm *svm = to_svm(vcpu);

1945 1946
	svm->vmcb->save.gdtr.limit = dt->size;
	svm->vmcb->save.gdtr.base = dt->address ;
1947
	mark_dirty(svm->vmcb, VMCB_DT);
A
Avi Kivity 已提交
1948 1949
}

1950 1951 1952 1953
static void svm_decache_cr0_guest_bits(struct kvm_vcpu *vcpu)
{
}

1954 1955 1956 1957
static void svm_decache_cr3(struct kvm_vcpu *vcpu)
{
}

1958
static void svm_decache_cr4_guest_bits(struct kvm_vcpu *vcpu)
1959 1960 1961
{
}

A
Avi Kivity 已提交
1962 1963 1964 1965 1966
static void update_cr0_intercept(struct vcpu_svm *svm)
{
	ulong gcr0 = svm->vcpu.arch.cr0;
	u64 *hcr0 = &svm->vmcb->save.cr0;

1967 1968
	*hcr0 = (*hcr0 & ~SVM_CR0_SELECTIVE_MASK)
		| (gcr0 & SVM_CR0_SELECTIVE_MASK);
A
Avi Kivity 已提交
1969

1970
	mark_dirty(svm->vmcb, VMCB_CR);
A
Avi Kivity 已提交
1971

1972
	if (gcr0 == *hcr0) {
1973 1974
		clr_cr_intercept(svm, INTERCEPT_CR0_READ);
		clr_cr_intercept(svm, INTERCEPT_CR0_WRITE);
A
Avi Kivity 已提交
1975
	} else {
1976 1977
		set_cr_intercept(svm, INTERCEPT_CR0_READ);
		set_cr_intercept(svm, INTERCEPT_CR0_WRITE);
A
Avi Kivity 已提交
1978 1979 1980
	}
}

A
Avi Kivity 已提交
1981 1982
static void svm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
{
1983 1984
	struct vcpu_svm *svm = to_svm(vcpu);

1985
#ifdef CONFIG_X86_64
1986
	if (vcpu->arch.efer & EFER_LME) {
1987
		if (!is_paging(vcpu) && (cr0 & X86_CR0_PG)) {
1988
			vcpu->arch.efer |= EFER_LMA;
1989
			svm->vmcb->save.efer |= EFER_LMA | EFER_LME;
A
Avi Kivity 已提交
1990 1991
		}

M
Mike Day 已提交
1992
		if (is_paging(vcpu) && !(cr0 & X86_CR0_PG)) {
1993
			vcpu->arch.efer &= ~EFER_LMA;
1994
			svm->vmcb->save.efer &= ~(EFER_LMA | EFER_LME);
A
Avi Kivity 已提交
1995 1996 1997
		}
	}
#endif
1998
	vcpu->arch.cr0 = cr0;
1999 2000 2001

	if (!npt_enabled)
		cr0 |= X86_CR0_PG | X86_CR0_WP;
2002

2003 2004 2005 2006 2007 2008 2009
	/*
	 * re-enable caching here because the QEMU bios
	 * does not do it - this results in some delay at
	 * reboot
	 */
	if (kvm_check_has_quirk(vcpu->kvm, KVM_X86_QUIRK_CD_NW_CLEARED))
		cr0 &= ~(X86_CR0_CD | X86_CR0_NW);
2010
	svm->vmcb->save.cr0 = cr0;
2011
	mark_dirty(svm->vmcb, VMCB_CR);
A
Avi Kivity 已提交
2012
	update_cr0_intercept(svm);
A
Avi Kivity 已提交
2013 2014
}

2015
static int svm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
A
Avi Kivity 已提交
2016
{
2017
	unsigned long host_cr4_mce = cr4_read_shadow() & X86_CR4_MCE;
2018 2019
	unsigned long old_cr4 = to_svm(vcpu)->vmcb->save.cr4;

2020 2021 2022
	if (cr4 & X86_CR4_VMXE)
		return 1;

2023
	if (npt_enabled && ((old_cr4 ^ cr4) & X86_CR4_PGE))
2024
		svm_flush_tlb(vcpu);
2025

2026 2027 2028
	vcpu->arch.cr4 = cr4;
	if (!npt_enabled)
		cr4 |= X86_CR4_PAE;
2029
	cr4 |= host_cr4_mce;
2030
	to_svm(vcpu)->vmcb->save.cr4 = cr4;
2031
	mark_dirty(to_svm(vcpu)->vmcb, VMCB_CR);
2032
	return 0;
A
Avi Kivity 已提交
2033 2034 2035 2036 2037
}

static void svm_set_segment(struct kvm_vcpu *vcpu,
			    struct kvm_segment *var, int seg)
{
2038
	struct vcpu_svm *svm = to_svm(vcpu);
A
Avi Kivity 已提交
2039 2040 2041 2042 2043
	struct vmcb_seg *s = svm_seg(vcpu, seg);

	s->base = var->base;
	s->limit = var->limit;
	s->selector = var->selector;
2044 2045 2046 2047 2048 2049 2050 2051
	s->attrib = (var->type & SVM_SELECTOR_TYPE_MASK);
	s->attrib |= (var->s & 1) << SVM_SELECTOR_S_SHIFT;
	s->attrib |= (var->dpl & 3) << SVM_SELECTOR_DPL_SHIFT;
	s->attrib |= ((var->present & 1) && !var->unusable) << SVM_SELECTOR_P_SHIFT;
	s->attrib |= (var->avl & 1) << SVM_SELECTOR_AVL_SHIFT;
	s->attrib |= (var->l & 1) << SVM_SELECTOR_L_SHIFT;
	s->attrib |= (var->db & 1) << SVM_SELECTOR_DB_SHIFT;
	s->attrib |= (var->g & 1) << SVM_SELECTOR_G_SHIFT;
P
Paolo Bonzini 已提交
2052 2053 2054 2055 2056 2057 2058 2059

	/*
	 * This is always accurate, except if SYSRET returned to a segment
	 * with SS.DPL != 3.  Intel does not have this quirk, and always
	 * forces SS.DPL to 3 on sysret, so we ignore that case; fixing it
	 * would entail passing the CPL to userspace and back.
	 */
	if (seg == VCPU_SREG_SS)
2060 2061
		/* This is symmetric with svm_get_segment() */
		svm->vmcb->save.cpl = (var->dpl & 3);
A
Avi Kivity 已提交
2062

2063
	mark_dirty(svm->vmcb, VMCB_SEG);
A
Avi Kivity 已提交
2064 2065
}

2066
static void update_bp_intercept(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
2067
{
J
Jan Kiszka 已提交
2068 2069
	struct vcpu_svm *svm = to_svm(vcpu);

2070
	clr_exception_intercept(svm, BP_VECTOR);
2071

J
Jan Kiszka 已提交
2072 2073
	if (vcpu->guest_debug & KVM_GUESTDBG_ENABLE) {
		if (vcpu->guest_debug & KVM_GUESTDBG_USE_SW_BP)
2074
			set_exception_intercept(svm, BP_VECTOR);
J
Jan Kiszka 已提交
2075 2076
	} else
		vcpu->guest_debug = 0;
2077 2078
}

2079
static void new_asid(struct vcpu_svm *svm, struct svm_cpu_data *sd)
A
Avi Kivity 已提交
2080
{
2081 2082 2083
	if (sd->next_asid > sd->max_asid) {
		++sd->asid_generation;
		sd->next_asid = 1;
2084
		svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ALL_ASID;
A
Avi Kivity 已提交
2085 2086
	}

2087 2088
	svm->asid_generation = sd->asid_generation;
	svm->vmcb->control.asid = sd->next_asid++;
2089 2090

	mark_dirty(svm->vmcb, VMCB_ASID);
A
Avi Kivity 已提交
2091 2092
}

J
Jan Kiszka 已提交
2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105
static u64 svm_get_dr6(struct kvm_vcpu *vcpu)
{
	return to_svm(vcpu)->vmcb->save.dr6;
}

static void svm_set_dr6(struct kvm_vcpu *vcpu, unsigned long value)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	svm->vmcb->save.dr6 = value;
	mark_dirty(svm->vmcb, VMCB_DR);
}

2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120
static void svm_sync_dirty_debug_regs(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	get_debugreg(vcpu->arch.db[0], 0);
	get_debugreg(vcpu->arch.db[1], 1);
	get_debugreg(vcpu->arch.db[2], 2);
	get_debugreg(vcpu->arch.db[3], 3);
	vcpu->arch.dr6 = svm_get_dr6(vcpu);
	vcpu->arch.dr7 = svm->vmcb->save.dr7;

	vcpu->arch.switch_db_regs &= ~KVM_DEBUGREG_WONT_EXIT;
	set_dr_intercepts(svm);
}

2121
static void svm_set_dr7(struct kvm_vcpu *vcpu, unsigned long value)
A
Avi Kivity 已提交
2122
{
2123 2124
	struct vcpu_svm *svm = to_svm(vcpu);

2125
	svm->vmcb->save.dr7 = value;
2126
	mark_dirty(svm->vmcb, VMCB_DR);
A
Avi Kivity 已提交
2127 2128
}

A
Avi Kivity 已提交
2129
static int pf_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2130
{
G
Gleb Natapov 已提交
2131
	u64 fault_address = svm->vmcb->control.exit_info_2;
2132
	u64 error_code = svm->vmcb->control.exit_info_1;
A
Avi Kivity 已提交
2133

2134
	return kvm_handle_page_fault(&svm->vcpu, error_code, fault_address,
2135
			svm->vmcb->control.insn_bytes,
2136
			svm->vmcb->control.insn_len, !npt_enabled);
A
Avi Kivity 已提交
2137 2138
}

A
Avi Kivity 已提交
2139
static int db_interception(struct vcpu_svm *svm)
J
Jan Kiszka 已提交
2140
{
A
Avi Kivity 已提交
2141 2142
	struct kvm_run *kvm_run = svm->vcpu.run;

J
Jan Kiszka 已提交
2143
	if (!(svm->vcpu.guest_debug &
2144
	      (KVM_GUESTDBG_SINGLESTEP | KVM_GUESTDBG_USE_HW_BP)) &&
J
Jan Kiszka 已提交
2145
		!svm->nmi_singlestep) {
J
Jan Kiszka 已提交
2146 2147 2148
		kvm_queue_exception(&svm->vcpu, DB_VECTOR);
		return 1;
	}
2149

J
Jan Kiszka 已提交
2150
	if (svm->nmi_singlestep) {
2151
		disable_nmi_singlestep(svm);
2152 2153 2154
	}

	if (svm->vcpu.guest_debug &
J
Joerg Roedel 已提交
2155
	    (KVM_GUESTDBG_SINGLESTEP | KVM_GUESTDBG_USE_HW_BP)) {
2156 2157 2158 2159 2160 2161 2162 2163
		kvm_run->exit_reason = KVM_EXIT_DEBUG;
		kvm_run->debug.arch.pc =
			svm->vmcb->save.cs.base + svm->vmcb->save.rip;
		kvm_run->debug.arch.exception = DB_VECTOR;
		return 0;
	}

	return 1;
J
Jan Kiszka 已提交
2164 2165
}

A
Avi Kivity 已提交
2166
static int bp_interception(struct vcpu_svm *svm)
J
Jan Kiszka 已提交
2167
{
A
Avi Kivity 已提交
2168 2169
	struct kvm_run *kvm_run = svm->vcpu.run;

J
Jan Kiszka 已提交
2170 2171 2172 2173 2174 2175
	kvm_run->exit_reason = KVM_EXIT_DEBUG;
	kvm_run->debug.arch.pc = svm->vmcb->save.cs.base + svm->vmcb->save.rip;
	kvm_run->debug.arch.exception = BP_VECTOR;
	return 0;
}

A
Avi Kivity 已提交
2176
static int ud_interception(struct vcpu_svm *svm)
2177 2178 2179
{
	int er;

2180
	er = emulate_instruction(&svm->vcpu, EMULTYPE_TRAP_UD);
2181
	if (er != EMULATE_DONE)
2182
		kvm_queue_exception(&svm->vcpu, UD_VECTOR);
2183 2184 2185
	return 1;
}

2186 2187 2188 2189 2190 2191
static int ac_interception(struct vcpu_svm *svm)
{
	kvm_queue_exception_e(&svm->vcpu, AC_VECTOR, 0);
	return 1;
}

2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230
static bool is_erratum_383(void)
{
	int err, i;
	u64 value;

	if (!erratum_383_found)
		return false;

	value = native_read_msr_safe(MSR_IA32_MC0_STATUS, &err);
	if (err)
		return false;

	/* Bit 62 may or may not be set for this mce */
	value &= ~(1ULL << 62);

	if (value != 0xb600000000010015ULL)
		return false;

	/* Clear MCi_STATUS registers */
	for (i = 0; i < 6; ++i)
		native_write_msr_safe(MSR_IA32_MCx_STATUS(i), 0, 0);

	value = native_read_msr_safe(MSR_IA32_MCG_STATUS, &err);
	if (!err) {
		u32 low, high;

		value &= ~(1ULL << 2);
		low    = lower_32_bits(value);
		high   = upper_32_bits(value);

		native_write_msr_safe(MSR_IA32_MCG_STATUS, low, high);
	}

	/* Flush tlb to evict multi-match entries */
	__flush_tlb_all();

	return true;
}

2231
static void svm_handle_mce(struct vcpu_svm *svm)
2232
{
2233 2234 2235 2236 2237 2238 2239
	if (is_erratum_383()) {
		/*
		 * Erratum 383 triggered. Guest state is corrupt so kill the
		 * guest.
		 */
		pr_err("KVM: Guest triggered AMD Erratum 383\n");

2240
		kvm_make_request(KVM_REQ_TRIPLE_FAULT, &svm->vcpu);
2241 2242 2243 2244

		return;
	}

2245 2246 2247 2248 2249 2250 2251 2252
	/*
	 * On an #MC intercept the MCE handler is not called automatically in
	 * the host. So do it by hand here.
	 */
	asm volatile (
		"int $0x12\n");
	/* not sure if we ever come back to this point */

2253 2254 2255 2256 2257
	return;
}

static int mc_interception(struct vcpu_svm *svm)
{
2258 2259 2260
	return 1;
}

A
Avi Kivity 已提交
2261
static int shutdown_interception(struct vcpu_svm *svm)
2262
{
A
Avi Kivity 已提交
2263 2264
	struct kvm_run *kvm_run = svm->vcpu.run;

2265 2266 2267 2268
	/*
	 * VMCB is undefined after a SHUTDOWN intercept
	 * so reinitialize it.
	 */
2269
	clear_page(svm->vmcb);
P
Paolo Bonzini 已提交
2270
	init_vmcb(svm);
2271 2272 2273 2274 2275

	kvm_run->exit_reason = KVM_EXIT_SHUTDOWN;
	return 0;
}

A
Avi Kivity 已提交
2276
static int io_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2277
{
2278
	struct kvm_vcpu *vcpu = &svm->vcpu;
M
Mike Day 已提交
2279
	u32 io_info = svm->vmcb->control.exit_info_1; /* address size bug? */
2280
	int size, in, string, ret;
2281
	unsigned port;
A
Avi Kivity 已提交
2282

R
Rusty Russell 已提交
2283
	++svm->vcpu.stat.io_exits;
2284
	string = (io_info & SVM_IOIO_STR_MASK) != 0;
2285
	in = (io_info & SVM_IOIO_TYPE_MASK) != 0;
2286
	if (string)
2287
		return emulate_instruction(vcpu, 0) == EMULATE_DONE;
2288

2289 2290
	port = io_info >> 16;
	size = (io_info & SVM_IOIO_SIZE_MASK) >> SVM_IOIO_SIZE_SHIFT;
2291
	svm->next_rip = svm->vmcb->control.exit_info_2;
2292
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
2293

2294 2295 2296 2297 2298 2299 2300 2301
	/*
	 * TODO: we might be squashing a KVM_GUESTDBG_SINGLESTEP-triggered
	 * KVM_EXIT_DEBUG here.
	 */
	if (in)
		return kvm_fast_pio_in(vcpu, size, port) && ret;
	else
		return kvm_fast_pio_out(vcpu, size, port) && ret;
A
Avi Kivity 已提交
2302 2303
}

A
Avi Kivity 已提交
2304
static int nmi_interception(struct vcpu_svm *svm)
2305 2306 2307 2308
{
	return 1;
}

A
Avi Kivity 已提交
2309
static int intr_interception(struct vcpu_svm *svm)
2310 2311 2312 2313 2314
{
	++svm->vcpu.stat.irq_exits;
	return 1;
}

A
Avi Kivity 已提交
2315
static int nop_on_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2316 2317 2318 2319
{
	return 1;
}

A
Avi Kivity 已提交
2320
static int halt_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2321
{
2322
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 1;
R
Rusty Russell 已提交
2323
	return kvm_emulate_halt(&svm->vcpu);
A
Avi Kivity 已提交
2324 2325
}

A
Avi Kivity 已提交
2326
static int vmmcall_interception(struct vcpu_svm *svm)
2327
{
2328
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
2329
	return kvm_emulate_hypercall(&svm->vcpu);
2330 2331
}

2332 2333 2334 2335 2336 2337 2338
static unsigned long nested_svm_get_tdp_cr3(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	return svm->nested.nested_cr3;
}

2339 2340 2341 2342 2343 2344 2345
static u64 nested_svm_get_tdp_pdptr(struct kvm_vcpu *vcpu, int index)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u64 cr3 = svm->nested.nested_cr3;
	u64 pdpte;
	int ret;

2346
	ret = kvm_vcpu_read_guest_page(vcpu, gpa_to_gfn(__sme_clr(cr3)), &pdpte,
2347
				       offset_in_page(cr3) + index * 8, 8);
2348 2349 2350 2351 2352
	if (ret)
		return 0;
	return pdpte;
}

2353 2354 2355 2356 2357
static void nested_svm_set_tdp_cr3(struct kvm_vcpu *vcpu,
				   unsigned long root)
{
	struct vcpu_svm *svm = to_svm(vcpu);

2358
	svm->vmcb->control.nested_cr3 = __sme_set(root);
2359
	mark_dirty(svm->vmcb, VMCB_NPT);
2360
	svm_flush_tlb(vcpu);
2361 2362
}

2363 2364
static void nested_svm_inject_npf_exit(struct kvm_vcpu *vcpu,
				       struct x86_exception *fault)
2365 2366 2367
{
	struct vcpu_svm *svm = to_svm(vcpu);

2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387
	if (svm->vmcb->control.exit_code != SVM_EXIT_NPF) {
		/*
		 * TODO: track the cause of the nested page fault, and
		 * correctly fill in the high bits of exit_info_1.
		 */
		svm->vmcb->control.exit_code = SVM_EXIT_NPF;
		svm->vmcb->control.exit_code_hi = 0;
		svm->vmcb->control.exit_info_1 = (1ULL << 32);
		svm->vmcb->control.exit_info_2 = fault->address;
	}

	svm->vmcb->control.exit_info_1 &= ~0xffffffffULL;
	svm->vmcb->control.exit_info_1 |= fault->error_code;

	/*
	 * The present bit is always zero for page structure faults on real
	 * hardware.
	 */
	if (svm->vmcb->control.exit_info_1 & (2ULL << 32))
		svm->vmcb->control.exit_info_1 &= ~1;
2388 2389 2390 2391

	nested_svm_vmexit(svm);
}

2392
static void nested_svm_init_mmu_context(struct kvm_vcpu *vcpu)
2393
{
2394 2395
	WARN_ON(mmu_is_nested(vcpu));
	kvm_init_shadow_mmu(vcpu);
2396 2397
	vcpu->arch.mmu.set_cr3           = nested_svm_set_tdp_cr3;
	vcpu->arch.mmu.get_cr3           = nested_svm_get_tdp_cr3;
2398
	vcpu->arch.mmu.get_pdptr         = nested_svm_get_tdp_pdptr;
2399
	vcpu->arch.mmu.inject_page_fault = nested_svm_inject_npf_exit;
2400
	vcpu->arch.mmu.shadow_root_level = get_npt_level(vcpu);
2401
	reset_shadow_zero_bits_mask(vcpu, &vcpu->arch.mmu);
2402 2403 2404 2405 2406 2407 2408 2409
	vcpu->arch.walk_mmu              = &vcpu->arch.nested_mmu;
}

static void nested_svm_uninit_mmu_context(struct kvm_vcpu *vcpu)
{
	vcpu->arch.walk_mmu = &vcpu->arch.mmu;
}

2410 2411
static int nested_svm_check_permissions(struct vcpu_svm *svm)
{
2412 2413
	if (!(svm->vcpu.arch.efer & EFER_SVME) ||
	    !is_paging(&svm->vcpu)) {
2414 2415 2416 2417 2418 2419 2420 2421 2422
		kvm_queue_exception(&svm->vcpu, UD_VECTOR);
		return 1;
	}

	if (svm->vmcb->save.cpl) {
		kvm_inject_gp(&svm->vcpu, 0);
		return 1;
	}

2423
	return 0;
2424 2425
}

2426 2427 2428
static int nested_svm_check_exception(struct vcpu_svm *svm, unsigned nr,
				      bool has_error_code, u32 error_code)
{
2429 2430
	int vmexit;

2431
	if (!is_guest_mode(&svm->vcpu))
2432
		return 0;
2433

2434 2435 2436 2437
	vmexit = nested_svm_intercept(svm);
	if (vmexit != NESTED_EXIT_DONE)
		return 0;

2438 2439 2440
	svm->vmcb->control.exit_code = SVM_EXIT_EXCP_BASE + nr;
	svm->vmcb->control.exit_code_hi = 0;
	svm->vmcb->control.exit_info_1 = error_code;
2441 2442 2443 2444 2445 2446 2447 2448 2449 2450

	/*
	 * FIXME: we should not write CR2 when L1 intercepts an L2 #PF exception.
	 * The fix is to add the ancillary datum (CR2 or DR6) to structs
	 * kvm_queued_exception and kvm_vcpu_events, so that CR2 and DR6 can be
	 * written only when inject_pending_event runs (DR6 would written here
	 * too).  This should be conditional on a new capability---if the
	 * capability is disabled, kvm_multiple_exception would write the
	 * ancillary information to CR2 or DR6, for backwards ABI-compatibility.
	 */
2451 2452 2453 2454
	if (svm->vcpu.arch.exception.nested_apf)
		svm->vmcb->control.exit_info_2 = svm->vcpu.arch.apf.nested_apf_token;
	else
		svm->vmcb->control.exit_info_2 = svm->vcpu.arch.cr2;
2455

2456
	svm->nested.exit_required = true;
2457
	return vmexit;
2458 2459
}

2460 2461
/* This function returns true if it is save to enable the irq window */
static inline bool nested_svm_intr(struct vcpu_svm *svm)
2462
{
2463
	if (!is_guest_mode(&svm->vcpu))
2464
		return true;
2465

2466
	if (!(svm->vcpu.arch.hflags & HF_VINTR_MASK))
2467
		return true;
2468

2469
	if (!(svm->vcpu.arch.hflags & HF_HIF_MASK))
2470
		return false;
2471

2472 2473 2474 2475 2476 2477 2478 2479
	/*
	 * if vmexit was already requested (by intercepted exception
	 * for instance) do not overwrite it with "external interrupt"
	 * vmexit.
	 */
	if (svm->nested.exit_required)
		return false;

2480 2481 2482
	svm->vmcb->control.exit_code   = SVM_EXIT_INTR;
	svm->vmcb->control.exit_info_1 = 0;
	svm->vmcb->control.exit_info_2 = 0;
2483

2484 2485 2486
	if (svm->nested.intercept & 1ULL) {
		/*
		 * The #vmexit can't be emulated here directly because this
G
Guo Chao 已提交
2487
		 * code path runs with irqs and preemption disabled. A
2488 2489 2490 2491
		 * #vmexit emulation might sleep. Only signal request for
		 * the #vmexit here.
		 */
		svm->nested.exit_required = true;
2492
		trace_kvm_nested_intr_vmexit(svm->vmcb->save.rip);
2493
		return false;
2494 2495
	}

2496
	return true;
2497 2498
}

2499 2500 2501
/* This function returns true if it is save to enable the nmi window */
static inline bool nested_svm_nmi(struct vcpu_svm *svm)
{
2502
	if (!is_guest_mode(&svm->vcpu))
2503 2504 2505 2506 2507 2508 2509 2510 2511
		return true;

	if (!(svm->nested.intercept & (1ULL << INTERCEPT_NMI)))
		return true;

	svm->vmcb->control.exit_code = SVM_EXIT_NMI;
	svm->nested.exit_required = true;

	return false;
2512 2513
}

2514
static void *nested_svm_map(struct vcpu_svm *svm, u64 gpa, struct page **_page)
2515 2516 2517
{
	struct page *page;

2518 2519
	might_sleep();

2520
	page = kvm_vcpu_gfn_to_page(&svm->vcpu, gpa >> PAGE_SHIFT);
2521 2522 2523
	if (is_error_page(page))
		goto error;

2524 2525 2526
	*_page = page;

	return kmap(page);
2527 2528 2529 2530 2531 2532 2533

error:
	kvm_inject_gp(&svm->vcpu, 0);

	return NULL;
}

2534
static void nested_svm_unmap(struct page *page)
2535
{
2536
	kunmap(page);
2537 2538 2539
	kvm_release_page_dirty(page);
}

2540 2541
static int nested_svm_intercept_ioio(struct vcpu_svm *svm)
{
2542 2543 2544
	unsigned port, size, iopm_len;
	u16 val, mask;
	u8 start_bit;
2545
	u64 gpa;
2546

2547 2548
	if (!(svm->nested.intercept & (1ULL << INTERCEPT_IOIO_PROT)))
		return NESTED_EXIT_HOST;
2549

2550
	port = svm->vmcb->control.exit_info_1 >> 16;
2551 2552
	size = (svm->vmcb->control.exit_info_1 & SVM_IOIO_SIZE_MASK) >>
		SVM_IOIO_SIZE_SHIFT;
2553
	gpa  = svm->nested.vmcb_iopm + (port / 8);
2554 2555 2556 2557
	start_bit = port % 8;
	iopm_len = (start_bit + size > 8) ? 2 : 1;
	mask = (0xf >> (4 - size)) << start_bit;
	val = 0;
2558

2559
	if (kvm_vcpu_read_guest(&svm->vcpu, gpa, &val, iopm_len))
2560
		return NESTED_EXIT_DONE;
2561

2562
	return (val & mask) ? NESTED_EXIT_DONE : NESTED_EXIT_HOST;
2563 2564
}

2565
static int nested_svm_exit_handled_msr(struct vcpu_svm *svm)
2566
{
2567 2568
	u32 offset, msr, value;
	int write, mask;
2569

2570
	if (!(svm->nested.intercept & (1ULL << INTERCEPT_MSR_PROT)))
2571
		return NESTED_EXIT_HOST;
2572

2573 2574 2575 2576
	msr    = svm->vcpu.arch.regs[VCPU_REGS_RCX];
	offset = svm_msrpm_offset(msr);
	write  = svm->vmcb->control.exit_info_1 & 1;
	mask   = 1 << ((2 * (msr & 0xf)) + write);
2577

2578 2579
	if (offset == MSR_INVALID)
		return NESTED_EXIT_DONE;
2580

2581 2582
	/* Offset is in 32 bit units but need in 8 bit units */
	offset *= 4;
2583

2584
	if (kvm_vcpu_read_guest(&svm->vcpu, svm->nested.vmcb_msrpm + offset, &value, 4))
2585
		return NESTED_EXIT_DONE;
2586

2587
	return (value & mask) ? NESTED_EXIT_DONE : NESTED_EXIT_HOST;
2588 2589
}

2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614
/* DB exceptions for our internal use must not cause vmexit */
static int nested_svm_intercept_db(struct vcpu_svm *svm)
{
	unsigned long dr6;

	/* if we're not singlestepping, it's not ours */
	if (!svm->nmi_singlestep)
		return NESTED_EXIT_DONE;

	/* if it's not a singlestep exception, it's not ours */
	if (kvm_get_dr(&svm->vcpu, 6, &dr6))
		return NESTED_EXIT_DONE;
	if (!(dr6 & DR6_BS))
		return NESTED_EXIT_DONE;

	/* if the guest is singlestepping, it should get the vmexit */
	if (svm->nmi_singlestep_guest_rflags & X86_EFLAGS_TF) {
		disable_nmi_singlestep(svm);
		return NESTED_EXIT_DONE;
	}

	/* it's ours, the nested hypervisor must not see this one */
	return NESTED_EXIT_HOST;
}

2615
static int nested_svm_exit_special(struct vcpu_svm *svm)
2616 2617
{
	u32 exit_code = svm->vmcb->control.exit_code;
2618

2619 2620 2621
	switch (exit_code) {
	case SVM_EXIT_INTR:
	case SVM_EXIT_NMI:
2622
	case SVM_EXIT_EXCP_BASE + MC_VECTOR:
2623 2624
		return NESTED_EXIT_HOST;
	case SVM_EXIT_NPF:
J
Joerg Roedel 已提交
2625
		/* For now we are always handling NPFs when using them */
2626 2627 2628 2629
		if (npt_enabled)
			return NESTED_EXIT_HOST;
		break;
	case SVM_EXIT_EXCP_BASE + PF_VECTOR:
G
Gleb Natapov 已提交
2630
		/* When we're shadowing, trap PFs, but not async PF */
2631
		if (!npt_enabled && svm->vcpu.arch.apf.host_apf_reason == 0)
2632 2633 2634 2635
			return NESTED_EXIT_HOST;
		break;
	default:
		break;
2636 2637
	}

2638 2639 2640 2641 2642 2643
	return NESTED_EXIT_CONTINUE;
}

/*
 * If this function returns true, this #vmexit was already handled
 */
2644
static int nested_svm_intercept(struct vcpu_svm *svm)
2645 2646 2647 2648
{
	u32 exit_code = svm->vmcb->control.exit_code;
	int vmexit = NESTED_EXIT_HOST;

2649
	switch (exit_code) {
2650
	case SVM_EXIT_MSR:
2651
		vmexit = nested_svm_exit_handled_msr(svm);
2652
		break;
2653 2654 2655
	case SVM_EXIT_IOIO:
		vmexit = nested_svm_intercept_ioio(svm);
		break;
2656 2657 2658
	case SVM_EXIT_READ_CR0 ... SVM_EXIT_WRITE_CR8: {
		u32 bit = 1U << (exit_code - SVM_EXIT_READ_CR0);
		if (svm->nested.intercept_cr & bit)
2659
			vmexit = NESTED_EXIT_DONE;
2660 2661
		break;
	}
2662 2663 2664
	case SVM_EXIT_READ_DR0 ... SVM_EXIT_WRITE_DR7: {
		u32 bit = 1U << (exit_code - SVM_EXIT_READ_DR0);
		if (svm->nested.intercept_dr & bit)
2665
			vmexit = NESTED_EXIT_DONE;
2666 2667 2668 2669
		break;
	}
	case SVM_EXIT_EXCP_BASE ... SVM_EXIT_EXCP_BASE + 0x1f: {
		u32 excp_bits = 1 << (exit_code - SVM_EXIT_EXCP_BASE);
2670 2671 2672 2673 2674 2675
		if (svm->nested.intercept_exceptions & excp_bits) {
			if (exit_code == SVM_EXIT_EXCP_BASE + DB_VECTOR)
				vmexit = nested_svm_intercept_db(svm);
			else
				vmexit = NESTED_EXIT_DONE;
		}
G
Gleb Natapov 已提交
2676 2677
		/* async page fault always cause vmexit */
		else if ((exit_code == SVM_EXIT_EXCP_BASE + PF_VECTOR) &&
2678
			 svm->vcpu.arch.exception.nested_apf != 0)
G
Gleb Natapov 已提交
2679
			vmexit = NESTED_EXIT_DONE;
2680 2681
		break;
	}
2682 2683 2684 2685
	case SVM_EXIT_ERR: {
		vmexit = NESTED_EXIT_DONE;
		break;
	}
2686 2687
	default: {
		u64 exit_bits = 1ULL << (exit_code - SVM_EXIT_INTR);
J
Joerg Roedel 已提交
2688
		if (svm->nested.intercept & exit_bits)
2689
			vmexit = NESTED_EXIT_DONE;
2690 2691 2692
	}
	}

2693 2694 2695 2696 2697 2698 2699 2700 2701 2702
	return vmexit;
}

static int nested_svm_exit_handled(struct vcpu_svm *svm)
{
	int vmexit;

	vmexit = nested_svm_intercept(svm);

	if (vmexit == NESTED_EXIT_DONE)
2703 2704 2705
		nested_svm_vmexit(svm);

	return vmexit;
2706 2707
}

2708 2709 2710 2711 2712
static inline void copy_vmcb_control_area(struct vmcb *dst_vmcb, struct vmcb *from_vmcb)
{
	struct vmcb_control_area *dst  = &dst_vmcb->control;
	struct vmcb_control_area *from = &from_vmcb->control;

2713
	dst->intercept_cr         = from->intercept_cr;
2714
	dst->intercept_dr         = from->intercept_dr;
2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734
	dst->intercept_exceptions = from->intercept_exceptions;
	dst->intercept            = from->intercept;
	dst->iopm_base_pa         = from->iopm_base_pa;
	dst->msrpm_base_pa        = from->msrpm_base_pa;
	dst->tsc_offset           = from->tsc_offset;
	dst->asid                 = from->asid;
	dst->tlb_ctl              = from->tlb_ctl;
	dst->int_ctl              = from->int_ctl;
	dst->int_vector           = from->int_vector;
	dst->int_state            = from->int_state;
	dst->exit_code            = from->exit_code;
	dst->exit_code_hi         = from->exit_code_hi;
	dst->exit_info_1          = from->exit_info_1;
	dst->exit_info_2          = from->exit_info_2;
	dst->exit_int_info        = from->exit_int_info;
	dst->exit_int_info_err    = from->exit_int_info_err;
	dst->nested_ctl           = from->nested_ctl;
	dst->event_inj            = from->event_inj;
	dst->event_inj_err        = from->event_inj_err;
	dst->nested_cr3           = from->nested_cr3;
2735
	dst->virt_ext              = from->virt_ext;
2736 2737
}

2738
static int nested_svm_vmexit(struct vcpu_svm *svm)
2739
{
2740
	struct vmcb *nested_vmcb;
2741
	struct vmcb *hsave = svm->nested.hsave;
J
Joerg Roedel 已提交
2742
	struct vmcb *vmcb = svm->vmcb;
2743
	struct page *page;
2744

2745 2746 2747 2748
	trace_kvm_nested_vmexit_inject(vmcb->control.exit_code,
				       vmcb->control.exit_info_1,
				       vmcb->control.exit_info_2,
				       vmcb->control.exit_int_info,
2749 2750
				       vmcb->control.exit_int_info_err,
				       KVM_ISA_SVM);
2751

2752
	nested_vmcb = nested_svm_map(svm, svm->nested.vmcb, &page);
2753 2754 2755
	if (!nested_vmcb)
		return 1;

2756 2757
	/* Exit Guest-Mode */
	leave_guest_mode(&svm->vcpu);
2758 2759
	svm->nested.vmcb = 0;

2760
	/* Give the current vmcb to the guest */
J
Joerg Roedel 已提交
2761 2762 2763 2764 2765 2766 2767 2768
	disable_gif(svm);

	nested_vmcb->save.es     = vmcb->save.es;
	nested_vmcb->save.cs     = vmcb->save.cs;
	nested_vmcb->save.ss     = vmcb->save.ss;
	nested_vmcb->save.ds     = vmcb->save.ds;
	nested_vmcb->save.gdtr   = vmcb->save.gdtr;
	nested_vmcb->save.idtr   = vmcb->save.idtr;
2769
	nested_vmcb->save.efer   = svm->vcpu.arch.efer;
2770
	nested_vmcb->save.cr0    = kvm_read_cr0(&svm->vcpu);
2771
	nested_vmcb->save.cr3    = kvm_read_cr3(&svm->vcpu);
J
Joerg Roedel 已提交
2772
	nested_vmcb->save.cr2    = vmcb->save.cr2;
2773
	nested_vmcb->save.cr4    = svm->vcpu.arch.cr4;
2774
	nested_vmcb->save.rflags = kvm_get_rflags(&svm->vcpu);
J
Joerg Roedel 已提交
2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790
	nested_vmcb->save.rip    = vmcb->save.rip;
	nested_vmcb->save.rsp    = vmcb->save.rsp;
	nested_vmcb->save.rax    = vmcb->save.rax;
	nested_vmcb->save.dr7    = vmcb->save.dr7;
	nested_vmcb->save.dr6    = vmcb->save.dr6;
	nested_vmcb->save.cpl    = vmcb->save.cpl;

	nested_vmcb->control.int_ctl           = vmcb->control.int_ctl;
	nested_vmcb->control.int_vector        = vmcb->control.int_vector;
	nested_vmcb->control.int_state         = vmcb->control.int_state;
	nested_vmcb->control.exit_code         = vmcb->control.exit_code;
	nested_vmcb->control.exit_code_hi      = vmcb->control.exit_code_hi;
	nested_vmcb->control.exit_info_1       = vmcb->control.exit_info_1;
	nested_vmcb->control.exit_info_2       = vmcb->control.exit_info_2;
	nested_vmcb->control.exit_int_info     = vmcb->control.exit_int_info;
	nested_vmcb->control.exit_int_info_err = vmcb->control.exit_int_info_err;
2791 2792 2793

	if (svm->nrips_enabled)
		nested_vmcb->control.next_rip  = vmcb->control.next_rip;
2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809

	/*
	 * If we emulate a VMRUN/#VMEXIT in the same host #vmexit cycle we have
	 * to make sure that we do not lose injected events. So check event_inj
	 * here and copy it to exit_int_info if it is valid.
	 * Exit_int_info and event_inj can't be both valid because the case
	 * below only happens on a VMRUN instruction intercept which has
	 * no valid exit_int_info set.
	 */
	if (vmcb->control.event_inj & SVM_EVTINJ_VALID) {
		struct vmcb_control_area *nc = &nested_vmcb->control;

		nc->exit_int_info     = vmcb->control.event_inj;
		nc->exit_int_info_err = vmcb->control.event_inj_err;
	}

J
Joerg Roedel 已提交
2810 2811 2812
	nested_vmcb->control.tlb_ctl           = 0;
	nested_vmcb->control.event_inj         = 0;
	nested_vmcb->control.event_inj_err     = 0;
2813 2814 2815 2816 2817 2818

	/* We always set V_INTR_MASKING and remember the old value in hflags */
	if (!(svm->vcpu.arch.hflags & HF_VINTR_MASK))
		nested_vmcb->control.int_ctl &= ~V_INTR_MASKING_MASK;

	/* Restore the original control entries */
2819
	copy_vmcb_control_area(vmcb, hsave);
2820

2821 2822
	kvm_clear_exception_queue(&svm->vcpu);
	kvm_clear_interrupt_queue(&svm->vcpu);
2823

2824 2825
	svm->nested.nested_cr3 = 0;

2826 2827 2828 2829 2830 2831 2832
	/* Restore selected save entries */
	svm->vmcb->save.es = hsave->save.es;
	svm->vmcb->save.cs = hsave->save.cs;
	svm->vmcb->save.ss = hsave->save.ss;
	svm->vmcb->save.ds = hsave->save.ds;
	svm->vmcb->save.gdtr = hsave->save.gdtr;
	svm->vmcb->save.idtr = hsave->save.idtr;
2833
	kvm_set_rflags(&svm->vcpu, hsave->save.rflags);
2834 2835 2836 2837 2838 2839 2840
	svm_set_efer(&svm->vcpu, hsave->save.efer);
	svm_set_cr0(&svm->vcpu, hsave->save.cr0 | X86_CR0_PE);
	svm_set_cr4(&svm->vcpu, hsave->save.cr4);
	if (npt_enabled) {
		svm->vmcb->save.cr3 = hsave->save.cr3;
		svm->vcpu.arch.cr3 = hsave->save.cr3;
	} else {
2841
		(void)kvm_set_cr3(&svm->vcpu, hsave->save.cr3);
2842 2843 2844 2845 2846 2847 2848 2849
	}
	kvm_register_write(&svm->vcpu, VCPU_REGS_RAX, hsave->save.rax);
	kvm_register_write(&svm->vcpu, VCPU_REGS_RSP, hsave->save.rsp);
	kvm_register_write(&svm->vcpu, VCPU_REGS_RIP, hsave->save.rip);
	svm->vmcb->save.dr7 = 0;
	svm->vmcb->save.cpl = 0;
	svm->vmcb->control.exit_int_info = 0;

2850 2851
	mark_all_dirty(svm->vmcb);

2852
	nested_svm_unmap(page);
2853

2854
	nested_svm_uninit_mmu_context(&svm->vcpu);
2855 2856 2857 2858 2859
	kvm_mmu_reset_context(&svm->vcpu);
	kvm_mmu_load(&svm->vcpu);

	return 0;
}
A
Alexander Graf 已提交
2860

2861
static bool nested_svm_vmrun_msrpm(struct vcpu_svm *svm)
A
Alexander Graf 已提交
2862
{
2863 2864
	/*
	 * This function merges the msr permission bitmaps of kvm and the
G
Guo Chao 已提交
2865
	 * nested vmcb. It is optimized in that it only merges the parts where
2866 2867
	 * the kvm msr permission bitmap may contain zero bits
	 */
A
Alexander Graf 已提交
2868
	int i;
2869

2870 2871
	if (!(svm->nested.intercept & (1ULL << INTERCEPT_MSR_PROT)))
		return true;
2872

2873 2874 2875
	for (i = 0; i < MSRPM_OFFSETS; i++) {
		u32 value, p;
		u64 offset;
2876

2877 2878
		if (msrpm_offsets[i] == 0xffffffff)
			break;
A
Alexander Graf 已提交
2879

2880 2881
		p      = msrpm_offsets[i];
		offset = svm->nested.vmcb_msrpm + (p * 4);
2882

2883
		if (kvm_vcpu_read_guest(&svm->vcpu, offset, &value, 4))
2884 2885 2886 2887
			return false;

		svm->nested.msrpm[p] = svm->msrpm[p] | value;
	}
A
Alexander Graf 已提交
2888

2889
	svm->vmcb->control.msrpm_base_pa = __sme_set(__pa(svm->nested.msrpm));
2890 2891

	return true;
A
Alexander Graf 已提交
2892 2893
}

2894 2895 2896 2897 2898
static bool nested_vmcb_checks(struct vmcb *vmcb)
{
	if ((vmcb->control.intercept & (1ULL << INTERCEPT_VMRUN)) == 0)
		return false;

2899 2900 2901
	if (vmcb->control.asid == 0)
		return false;

2902 2903 2904
	if (vmcb->control.nested_ctl && !npt_enabled)
		return false;

2905 2906 2907
	return true;
}

2908
static bool nested_svm_vmrun(struct vcpu_svm *svm)
A
Alexander Graf 已提交
2909
{
2910
	struct vmcb *nested_vmcb;
2911
	struct vmcb *hsave = svm->nested.hsave;
J
Joerg Roedel 已提交
2912
	struct vmcb *vmcb = svm->vmcb;
2913
	struct page *page;
2914
	u64 vmcb_gpa;
A
Alexander Graf 已提交
2915

2916
	vmcb_gpa = svm->vmcb->save.rax;
A
Alexander Graf 已提交
2917

2918
	nested_vmcb = nested_svm_map(svm, svm->vmcb->save.rax, &page);
2919 2920 2921
	if (!nested_vmcb)
		return false;

2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932
	if (!nested_vmcb_checks(nested_vmcb)) {
		nested_vmcb->control.exit_code    = SVM_EXIT_ERR;
		nested_vmcb->control.exit_code_hi = 0;
		nested_vmcb->control.exit_info_1  = 0;
		nested_vmcb->control.exit_info_2  = 0;

		nested_svm_unmap(page);

		return false;
	}

2933
	trace_kvm_nested_vmrun(svm->vmcb->save.rip, vmcb_gpa,
2934 2935 2936 2937 2938
			       nested_vmcb->save.rip,
			       nested_vmcb->control.int_ctl,
			       nested_vmcb->control.event_inj,
			       nested_vmcb->control.nested_ctl);

2939 2940
	trace_kvm_nested_intercepts(nested_vmcb->control.intercept_cr & 0xffff,
				    nested_vmcb->control.intercept_cr >> 16,
2941 2942 2943
				    nested_vmcb->control.intercept_exceptions,
				    nested_vmcb->control.intercept);

A
Alexander Graf 已提交
2944
	/* Clear internal status */
2945 2946
	kvm_clear_exception_queue(&svm->vcpu);
	kvm_clear_interrupt_queue(&svm->vcpu);
A
Alexander Graf 已提交
2947

J
Joerg Roedel 已提交
2948 2949 2950 2951
	/*
	 * Save the old vmcb, so we don't need to pick what we save, but can
	 * restore everything when a VMEXIT occurs
	 */
J
Joerg Roedel 已提交
2952 2953 2954 2955 2956 2957
	hsave->save.es     = vmcb->save.es;
	hsave->save.cs     = vmcb->save.cs;
	hsave->save.ss     = vmcb->save.ss;
	hsave->save.ds     = vmcb->save.ds;
	hsave->save.gdtr   = vmcb->save.gdtr;
	hsave->save.idtr   = vmcb->save.idtr;
2958
	hsave->save.efer   = svm->vcpu.arch.efer;
2959
	hsave->save.cr0    = kvm_read_cr0(&svm->vcpu);
J
Joerg Roedel 已提交
2960
	hsave->save.cr4    = svm->vcpu.arch.cr4;
2961
	hsave->save.rflags = kvm_get_rflags(&svm->vcpu);
2962
	hsave->save.rip    = kvm_rip_read(&svm->vcpu);
J
Joerg Roedel 已提交
2963 2964 2965 2966 2967
	hsave->save.rsp    = vmcb->save.rsp;
	hsave->save.rax    = vmcb->save.rax;
	if (npt_enabled)
		hsave->save.cr3    = vmcb->save.cr3;
	else
2968
		hsave->save.cr3    = kvm_read_cr3(&svm->vcpu);
J
Joerg Roedel 已提交
2969

2970
	copy_vmcb_control_area(hsave, vmcb);
A
Alexander Graf 已提交
2971

2972
	if (kvm_get_rflags(&svm->vcpu) & X86_EFLAGS_IF)
A
Alexander Graf 已提交
2973 2974 2975 2976
		svm->vcpu.arch.hflags |= HF_HIF_MASK;
	else
		svm->vcpu.arch.hflags &= ~HF_HIF_MASK;

2977 2978 2979 2980 2981 2982
	if (nested_vmcb->control.nested_ctl) {
		kvm_mmu_unload(&svm->vcpu);
		svm->nested.nested_cr3 = nested_vmcb->control.nested_cr3;
		nested_svm_init_mmu_context(&svm->vcpu);
	}

A
Alexander Graf 已提交
2983 2984 2985 2986 2987 2988 2989
	/* Load the nested guest state */
	svm->vmcb->save.es = nested_vmcb->save.es;
	svm->vmcb->save.cs = nested_vmcb->save.cs;
	svm->vmcb->save.ss = nested_vmcb->save.ss;
	svm->vmcb->save.ds = nested_vmcb->save.ds;
	svm->vmcb->save.gdtr = nested_vmcb->save.gdtr;
	svm->vmcb->save.idtr = nested_vmcb->save.idtr;
2990
	kvm_set_rflags(&svm->vcpu, nested_vmcb->save.rflags);
A
Alexander Graf 已提交
2991 2992 2993 2994 2995 2996
	svm_set_efer(&svm->vcpu, nested_vmcb->save.efer);
	svm_set_cr0(&svm->vcpu, nested_vmcb->save.cr0);
	svm_set_cr4(&svm->vcpu, nested_vmcb->save.cr4);
	if (npt_enabled) {
		svm->vmcb->save.cr3 = nested_vmcb->save.cr3;
		svm->vcpu.arch.cr3 = nested_vmcb->save.cr3;
2997
	} else
2998
		(void)kvm_set_cr3(&svm->vcpu, nested_vmcb->save.cr3);
2999 3000 3001 3002

	/* Guest paging mode is active - reset mmu */
	kvm_mmu_reset_context(&svm->vcpu);

J
Joerg Roedel 已提交
3003
	svm->vmcb->save.cr2 = svm->vcpu.arch.cr2 = nested_vmcb->save.cr2;
A
Alexander Graf 已提交
3004 3005 3006
	kvm_register_write(&svm->vcpu, VCPU_REGS_RAX, nested_vmcb->save.rax);
	kvm_register_write(&svm->vcpu, VCPU_REGS_RSP, nested_vmcb->save.rsp);
	kvm_register_write(&svm->vcpu, VCPU_REGS_RIP, nested_vmcb->save.rip);
J
Joerg Roedel 已提交
3007

A
Alexander Graf 已提交
3008 3009 3010 3011 3012 3013 3014 3015
	/* In case we don't even reach vcpu_run, the fields are not updated */
	svm->vmcb->save.rax = nested_vmcb->save.rax;
	svm->vmcb->save.rsp = nested_vmcb->save.rsp;
	svm->vmcb->save.rip = nested_vmcb->save.rip;
	svm->vmcb->save.dr7 = nested_vmcb->save.dr7;
	svm->vmcb->save.dr6 = nested_vmcb->save.dr6;
	svm->vmcb->save.cpl = nested_vmcb->save.cpl;

3016
	svm->nested.vmcb_msrpm = nested_vmcb->control.msrpm_base_pa & ~0x0fffULL;
3017
	svm->nested.vmcb_iopm  = nested_vmcb->control.iopm_base_pa  & ~0x0fffULL;
A
Alexander Graf 已提交
3018

J
Joerg Roedel 已提交
3019
	/* cache intercepts */
3020
	svm->nested.intercept_cr         = nested_vmcb->control.intercept_cr;
3021
	svm->nested.intercept_dr         = nested_vmcb->control.intercept_dr;
J
Joerg Roedel 已提交
3022 3023 3024
	svm->nested.intercept_exceptions = nested_vmcb->control.intercept_exceptions;
	svm->nested.intercept            = nested_vmcb->control.intercept;

3025
	svm_flush_tlb(&svm->vcpu);
A
Alexander Graf 已提交
3026 3027 3028 3029 3030 3031
	svm->vmcb->control.int_ctl = nested_vmcb->control.int_ctl | V_INTR_MASKING_MASK;
	if (nested_vmcb->control.int_ctl & V_INTR_MASKING_MASK)
		svm->vcpu.arch.hflags |= HF_VINTR_MASK;
	else
		svm->vcpu.arch.hflags &= ~HF_VINTR_MASK;

3032 3033
	if (svm->vcpu.arch.hflags & HF_VINTR_MASK) {
		/* We only want the cr8 intercept bits of the guest */
3034 3035
		clr_cr_intercept(svm, INTERCEPT_CR8_READ);
		clr_cr_intercept(svm, INTERCEPT_CR8_WRITE);
3036 3037
	}

3038
	/* We don't want to see VMMCALLs from a nested guest */
3039
	clr_intercept(svm, INTERCEPT_VMMCALL);
3040

3041
	svm->vmcb->control.virt_ext = nested_vmcb->control.virt_ext;
A
Alexander Graf 已提交
3042 3043 3044 3045 3046 3047
	svm->vmcb->control.int_vector = nested_vmcb->control.int_vector;
	svm->vmcb->control.int_state = nested_vmcb->control.int_state;
	svm->vmcb->control.tsc_offset += nested_vmcb->control.tsc_offset;
	svm->vmcb->control.event_inj = nested_vmcb->control.event_inj;
	svm->vmcb->control.event_inj_err = nested_vmcb->control.event_inj_err;

3048
	nested_svm_unmap(page);
3049

3050 3051 3052
	/* Enter Guest-Mode */
	enter_guest_mode(&svm->vcpu);

3053 3054 3055 3056 3057 3058
	/*
	 * Merge guest and host intercepts - must be called  with vcpu in
	 * guest-mode to take affect here
	 */
	recalc_intercepts(svm);

3059
	svm->nested.vmcb = vmcb_gpa;
3060

3061
	enable_gif(svm);
A
Alexander Graf 已提交
3062

3063 3064
	mark_all_dirty(svm->vmcb);

3065
	return true;
A
Alexander Graf 已提交
3066 3067
}

3068
static void nested_svm_vmloadsave(struct vmcb *from_vmcb, struct vmcb *to_vmcb)
3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083
{
	to_vmcb->save.fs = from_vmcb->save.fs;
	to_vmcb->save.gs = from_vmcb->save.gs;
	to_vmcb->save.tr = from_vmcb->save.tr;
	to_vmcb->save.ldtr = from_vmcb->save.ldtr;
	to_vmcb->save.kernel_gs_base = from_vmcb->save.kernel_gs_base;
	to_vmcb->save.star = from_vmcb->save.star;
	to_vmcb->save.lstar = from_vmcb->save.lstar;
	to_vmcb->save.cstar = from_vmcb->save.cstar;
	to_vmcb->save.sfmask = from_vmcb->save.sfmask;
	to_vmcb->save.sysenter_cs = from_vmcb->save.sysenter_cs;
	to_vmcb->save.sysenter_esp = from_vmcb->save.sysenter_esp;
	to_vmcb->save.sysenter_eip = from_vmcb->save.sysenter_eip;
}

A
Avi Kivity 已提交
3084
static int vmload_interception(struct vcpu_svm *svm)
3085
{
3086
	struct vmcb *nested_vmcb;
3087
	struct page *page;
3088
	int ret;
3089

3090 3091 3092
	if (nested_svm_check_permissions(svm))
		return 1;

3093
	nested_vmcb = nested_svm_map(svm, svm->vmcb->save.rax, &page);
3094 3095 3096
	if (!nested_vmcb)
		return 1;

3097
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3098
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3099

3100
	nested_svm_vmloadsave(nested_vmcb, svm->vmcb);
3101
	nested_svm_unmap(page);
3102

3103
	return ret;
3104 3105
}

A
Avi Kivity 已提交
3106
static int vmsave_interception(struct vcpu_svm *svm)
3107
{
3108
	struct vmcb *nested_vmcb;
3109
	struct page *page;
3110
	int ret;
3111

3112 3113 3114
	if (nested_svm_check_permissions(svm))
		return 1;

3115
	nested_vmcb = nested_svm_map(svm, svm->vmcb->save.rax, &page);
3116 3117 3118
	if (!nested_vmcb)
		return 1;

3119
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3120
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3121

3122
	nested_svm_vmloadsave(svm->vmcb, nested_vmcb);
3123
	nested_svm_unmap(page);
3124

3125
	return ret;
3126 3127
}

A
Avi Kivity 已提交
3128
static int vmrun_interception(struct vcpu_svm *svm)
A
Alexander Graf 已提交
3129 3130 3131 3132
{
	if (nested_svm_check_permissions(svm))
		return 1;

3133 3134
	/* Save rip after vmrun instruction */
	kvm_rip_write(&svm->vcpu, kvm_rip_read(&svm->vcpu) + 3);
A
Alexander Graf 已提交
3135

3136
	if (!nested_svm_vmrun(svm))
A
Alexander Graf 已提交
3137 3138
		return 1;

3139
	if (!nested_svm_vmrun_msrpm(svm))
3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151
		goto failed;

	return 1;

failed:

	svm->vmcb->control.exit_code    = SVM_EXIT_ERR;
	svm->vmcb->control.exit_code_hi = 0;
	svm->vmcb->control.exit_info_1  = 0;
	svm->vmcb->control.exit_info_2  = 0;

	nested_svm_vmexit(svm);
A
Alexander Graf 已提交
3152 3153 3154 3155

	return 1;
}

A
Avi Kivity 已提交
3156
static int stgi_interception(struct vcpu_svm *svm)
3157
{
3158 3159
	int ret;

3160 3161 3162
	if (nested_svm_check_permissions(svm))
		return 1;

3163 3164 3165 3166 3167 3168 3169
	/*
	 * If VGIF is enabled, the STGI intercept is only added to
	 * detect the opening of the NMI window; remove it now.
	 */
	if (vgif_enabled(svm))
		clr_intercept(svm, INTERCEPT_STGI);

3170
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3171
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3172
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
3173

3174
	enable_gif(svm);
3175

3176
	return ret;
3177 3178
}

A
Avi Kivity 已提交
3179
static int clgi_interception(struct vcpu_svm *svm)
3180
{
3181 3182
	int ret;

3183 3184 3185 3186
	if (nested_svm_check_permissions(svm))
		return 1;

	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3187
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3188

3189
	disable_gif(svm);
3190 3191

	/* After a CLGI no interrupts should come */
3192 3193 3194 3195 3196
	if (!kvm_vcpu_apicv_active(&svm->vcpu)) {
		svm_clear_vintr(svm);
		svm->vmcb->control.int_ctl &= ~V_IRQ_MASK;
		mark_dirty(svm->vmcb, VMCB_INTR);
	}
3197

3198
	return ret;
3199 3200
}

A
Avi Kivity 已提交
3201
static int invlpga_interception(struct vcpu_svm *svm)
A
Alexander Graf 已提交
3202 3203 3204
{
	struct kvm_vcpu *vcpu = &svm->vcpu;

3205 3206
	trace_kvm_invlpga(svm->vmcb->save.rip, kvm_register_read(&svm->vcpu, VCPU_REGS_RCX),
			  kvm_register_read(&svm->vcpu, VCPU_REGS_RAX));
3207

A
Alexander Graf 已提交
3208
	/* Let's treat INVLPGA the same as INVLPG (can be optimized!) */
3209
	kvm_mmu_invlpg(vcpu, kvm_register_read(&svm->vcpu, VCPU_REGS_RAX));
A
Alexander Graf 已提交
3210 3211

	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3212
	return kvm_skip_emulated_instruction(&svm->vcpu);
A
Alexander Graf 已提交
3213 3214
}

3215 3216
static int skinit_interception(struct vcpu_svm *svm)
{
3217
	trace_kvm_skinit(svm->vmcb->save.rip, kvm_register_read(&svm->vcpu, VCPU_REGS_RAX));
3218 3219 3220 3221 3222

	kvm_queue_exception(&svm->vcpu, UD_VECTOR);
	return 1;
}

D
David Kaplan 已提交
3223 3224
static int wbinvd_interception(struct vcpu_svm *svm)
{
3225
	return kvm_emulate_wbinvd(&svm->vcpu);
D
David Kaplan 已提交
3226 3227
}

J
Joerg Roedel 已提交
3228 3229 3230 3231 3232 3233 3234
static int xsetbv_interception(struct vcpu_svm *svm)
{
	u64 new_bv = kvm_read_edx_eax(&svm->vcpu);
	u32 index = kvm_register_read(&svm->vcpu, VCPU_REGS_RCX);

	if (kvm_set_xcr(&svm->vcpu, index, new_bv) == 0) {
		svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3235
		return kvm_skip_emulated_instruction(&svm->vcpu);
J
Joerg Roedel 已提交
3236 3237 3238 3239 3240
	}

	return 1;
}

A
Avi Kivity 已提交
3241
static int task_switch_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3242
{
3243
	u16 tss_selector;
3244 3245 3246
	int reason;
	int int_type = svm->vmcb->control.exit_int_info &
		SVM_EXITINTINFO_TYPE_MASK;
3247
	int int_vec = svm->vmcb->control.exit_int_info & SVM_EVTINJ_VEC_MASK;
3248 3249 3250 3251
	uint32_t type =
		svm->vmcb->control.exit_int_info & SVM_EXITINTINFO_TYPE_MASK;
	uint32_t idt_v =
		svm->vmcb->control.exit_int_info & SVM_EXITINTINFO_VALID;
3252 3253
	bool has_error_code = false;
	u32 error_code = 0;
3254 3255

	tss_selector = (u16)svm->vmcb->control.exit_info_1;
3256

3257 3258
	if (svm->vmcb->control.exit_info_2 &
	    (1ULL << SVM_EXITINFOSHIFT_TS_REASON_IRET))
3259 3260 3261 3262
		reason = TASK_SWITCH_IRET;
	else if (svm->vmcb->control.exit_info_2 &
		 (1ULL << SVM_EXITINFOSHIFT_TS_REASON_JMP))
		reason = TASK_SWITCH_JMP;
3263
	else if (idt_v)
3264 3265 3266 3267
		reason = TASK_SWITCH_GATE;
	else
		reason = TASK_SWITCH_CALL;

3268 3269 3270 3271 3272 3273
	if (reason == TASK_SWITCH_GATE) {
		switch (type) {
		case SVM_EXITINTINFO_TYPE_NMI:
			svm->vcpu.arch.nmi_injected = false;
			break;
		case SVM_EXITINTINFO_TYPE_EXEPT:
3274 3275 3276 3277 3278 3279
			if (svm->vmcb->control.exit_info_2 &
			    (1ULL << SVM_EXITINFOSHIFT_TS_HAS_ERROR_CODE)) {
				has_error_code = true;
				error_code =
					(u32)svm->vmcb->control.exit_info_2;
			}
3280 3281 3282 3283 3284 3285 3286 3287 3288
			kvm_clear_exception_queue(&svm->vcpu);
			break;
		case SVM_EXITINTINFO_TYPE_INTR:
			kvm_clear_interrupt_queue(&svm->vcpu);
			break;
		default:
			break;
		}
	}
3289

3290 3291 3292
	if (reason != TASK_SWITCH_GATE ||
	    int_type == SVM_EXITINTINFO_TYPE_SOFT ||
	    (int_type == SVM_EXITINTINFO_TYPE_EXEPT &&
3293 3294
	     (int_vec == OF_VECTOR || int_vec == BP_VECTOR)))
		skip_emulated_instruction(&svm->vcpu);
3295

3296 3297 3298 3299
	if (int_type != SVM_EXITINTINFO_TYPE_SOFT)
		int_vec = -1;

	if (kvm_task_switch(&svm->vcpu, tss_selector, int_vec, reason,
3300 3301 3302 3303 3304 3305 3306
				has_error_code, error_code) == EMULATE_FAIL) {
		svm->vcpu.run->exit_reason = KVM_EXIT_INTERNAL_ERROR;
		svm->vcpu.run->internal.suberror = KVM_INTERNAL_ERROR_EMULATION;
		svm->vcpu.run->internal.ndata = 0;
		return 0;
	}
	return 1;
A
Avi Kivity 已提交
3307 3308
}

A
Avi Kivity 已提交
3309
static int cpuid_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3310
{
3311
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 2;
3312
	return kvm_emulate_cpuid(&svm->vcpu);
A
Avi Kivity 已提交
3313 3314
}

A
Avi Kivity 已提交
3315
static int iret_interception(struct vcpu_svm *svm)
3316 3317
{
	++svm->vcpu.stat.nmi_window_exits;
3318
	clr_intercept(svm, INTERCEPT_IRET);
3319
	svm->vcpu.arch.hflags |= HF_IRET_MASK;
3320
	svm->nmi_iret_rip = kvm_rip_read(&svm->vcpu);
3321
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
3322 3323 3324
	return 1;
}

A
Avi Kivity 已提交
3325
static int invlpg_interception(struct vcpu_svm *svm)
M
Marcelo Tosatti 已提交
3326
{
3327 3328 3329 3330
	if (!static_cpu_has(X86_FEATURE_DECODEASSISTS))
		return emulate_instruction(&svm->vcpu, 0) == EMULATE_DONE;

	kvm_mmu_invlpg(&svm->vcpu, svm->vmcb->control.exit_info_1);
3331
	return kvm_skip_emulated_instruction(&svm->vcpu);
M
Marcelo Tosatti 已提交
3332 3333
}

A
Avi Kivity 已提交
3334
static int emulate_on_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3335
{
3336
	return emulate_instruction(&svm->vcpu, 0) == EMULATE_DONE;
A
Avi Kivity 已提交
3337 3338
}

A
Avi Kivity 已提交
3339 3340 3341 3342 3343 3344 3345 3346
static int rdpmc_interception(struct vcpu_svm *svm)
{
	int err;

	if (!static_cpu_has(X86_FEATURE_NRIPS))
		return emulate_on_interception(svm);

	err = kvm_rdpmc(&svm->vcpu);
3347
	return kvm_complete_insn_gp(&svm->vcpu, err);
A
Avi Kivity 已提交
3348 3349
}

3350 3351
static bool check_selective_cr0_intercepted(struct vcpu_svm *svm,
					    unsigned long val)
3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373
{
	unsigned long cr0 = svm->vcpu.arch.cr0;
	bool ret = false;
	u64 intercept;

	intercept = svm->nested.intercept;

	if (!is_guest_mode(&svm->vcpu) ||
	    (!(intercept & (1ULL << INTERCEPT_SELECTIVE_CR0))))
		return false;

	cr0 &= ~SVM_CR0_SELECTIVE_MASK;
	val &= ~SVM_CR0_SELECTIVE_MASK;

	if (cr0 ^ val) {
		svm->vmcb->control.exit_code = SVM_EXIT_CR0_SEL_WRITE;
		ret = (nested_svm_exit_handled(svm) == NESTED_EXIT_DONE);
	}

	return ret;
}

3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388
#define CR_VALID (1ULL << 63)

static int cr_interception(struct vcpu_svm *svm)
{
	int reg, cr;
	unsigned long val;
	int err;

	if (!static_cpu_has(X86_FEATURE_DECODEASSISTS))
		return emulate_on_interception(svm);

	if (unlikely((svm->vmcb->control.exit_info_1 & CR_VALID) == 0))
		return emulate_on_interception(svm);

	reg = svm->vmcb->control.exit_info_1 & SVM_EXITINFO_REG_MASK;
3389 3390 3391 3392
	if (svm->vmcb->control.exit_code == SVM_EXIT_CR0_SEL_WRITE)
		cr = SVM_EXIT_WRITE_CR0 - SVM_EXIT_READ_CR0;
	else
		cr = svm->vmcb->control.exit_code - SVM_EXIT_READ_CR0;
3393 3394 3395 3396 3397 3398 3399

	err = 0;
	if (cr >= 16) { /* mov to cr */
		cr -= 16;
		val = kvm_register_read(&svm->vcpu, reg);
		switch (cr) {
		case 0:
3400 3401
			if (!check_selective_cr0_intercepted(svm, val))
				err = kvm_set_cr0(&svm->vcpu, val);
3402 3403 3404
			else
				return 1;

3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428
			break;
		case 3:
			err = kvm_set_cr3(&svm->vcpu, val);
			break;
		case 4:
			err = kvm_set_cr4(&svm->vcpu, val);
			break;
		case 8:
			err = kvm_set_cr8(&svm->vcpu, val);
			break;
		default:
			WARN(1, "unhandled write to CR%d", cr);
			kvm_queue_exception(&svm->vcpu, UD_VECTOR);
			return 1;
		}
	} else { /* mov from cr */
		switch (cr) {
		case 0:
			val = kvm_read_cr0(&svm->vcpu);
			break;
		case 2:
			val = svm->vcpu.arch.cr2;
			break;
		case 3:
3429
			val = kvm_read_cr3(&svm->vcpu);
3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443
			break;
		case 4:
			val = kvm_read_cr4(&svm->vcpu);
			break;
		case 8:
			val = kvm_get_cr8(&svm->vcpu);
			break;
		default:
			WARN(1, "unhandled read from CR%d", cr);
			kvm_queue_exception(&svm->vcpu, UD_VECTOR);
			return 1;
		}
		kvm_register_write(&svm->vcpu, reg, val);
	}
3444
	return kvm_complete_insn_gp(&svm->vcpu, err);
3445 3446
}

3447 3448 3449 3450 3451
static int dr_interception(struct vcpu_svm *svm)
{
	int reg, dr;
	unsigned long val;

3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462
	if (svm->vcpu.guest_debug == 0) {
		/*
		 * No more DR vmexits; force a reload of the debug registers
		 * and reenter on this instruction.  The next vmexit will
		 * retrieve the full state of the debug registers.
		 */
		clr_dr_intercepts(svm);
		svm->vcpu.arch.switch_db_regs |= KVM_DEBUGREG_WONT_EXIT;
		return 1;
	}

3463 3464 3465 3466 3467 3468 3469
	if (!boot_cpu_has(X86_FEATURE_DECODEASSISTS))
		return emulate_on_interception(svm);

	reg = svm->vmcb->control.exit_info_1 & SVM_EXITINFO_REG_MASK;
	dr = svm->vmcb->control.exit_code - SVM_EXIT_READ_DR0;

	if (dr >= 16) { /* mov to DRn */
3470 3471
		if (!kvm_require_dr(&svm->vcpu, dr - 16))
			return 1;
3472 3473 3474
		val = kvm_register_read(&svm->vcpu, reg);
		kvm_set_dr(&svm->vcpu, dr - 16, val);
	} else {
3475 3476 3477 3478
		if (!kvm_require_dr(&svm->vcpu, dr))
			return 1;
		kvm_get_dr(&svm->vcpu, dr, &val);
		kvm_register_write(&svm->vcpu, reg, val);
3479 3480
	}

3481
	return kvm_skip_emulated_instruction(&svm->vcpu);
3482 3483
}

A
Avi Kivity 已提交
3484
static int cr8_write_interception(struct vcpu_svm *svm)
3485
{
A
Avi Kivity 已提交
3486
	struct kvm_run *kvm_run = svm->vcpu.run;
A
Andre Przywara 已提交
3487
	int r;
A
Avi Kivity 已提交
3488

3489 3490
	u8 cr8_prev = kvm_get_cr8(&svm->vcpu);
	/* instruction emulation calls kvm_set_cr8() */
3491
	r = cr_interception(svm);
3492
	if (lapic_in_kernel(&svm->vcpu))
3493
		return r;
3494
	if (cr8_prev <= kvm_get_cr8(&svm->vcpu))
3495
		return r;
3496 3497 3498 3499
	kvm_run->exit_reason = KVM_EXIT_SET_TPR;
	return 0;
}

3500
static int svm_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
A
Avi Kivity 已提交
3501
{
3502 3503
	struct vcpu_svm *svm = to_svm(vcpu);

3504
	switch (msr_info->index) {
3505
	case MSR_IA32_TSC: {
3506
		msr_info->data = svm->vmcb->control.tsc_offset +
3507
			kvm_scale_tsc(vcpu, rdtsc());
3508

A
Avi Kivity 已提交
3509 3510
		break;
	}
B
Brian Gerst 已提交
3511
	case MSR_STAR:
3512
		msr_info->data = svm->vmcb->save.star;
A
Avi Kivity 已提交
3513
		break;
3514
#ifdef CONFIG_X86_64
A
Avi Kivity 已提交
3515
	case MSR_LSTAR:
3516
		msr_info->data = svm->vmcb->save.lstar;
A
Avi Kivity 已提交
3517 3518
		break;
	case MSR_CSTAR:
3519
		msr_info->data = svm->vmcb->save.cstar;
A
Avi Kivity 已提交
3520 3521
		break;
	case MSR_KERNEL_GS_BASE:
3522
		msr_info->data = svm->vmcb->save.kernel_gs_base;
A
Avi Kivity 已提交
3523 3524
		break;
	case MSR_SYSCALL_MASK:
3525
		msr_info->data = svm->vmcb->save.sfmask;
A
Avi Kivity 已提交
3526 3527 3528
		break;
#endif
	case MSR_IA32_SYSENTER_CS:
3529
		msr_info->data = svm->vmcb->save.sysenter_cs;
A
Avi Kivity 已提交
3530 3531
		break;
	case MSR_IA32_SYSENTER_EIP:
3532
		msr_info->data = svm->sysenter_eip;
A
Avi Kivity 已提交
3533 3534
		break;
	case MSR_IA32_SYSENTER_ESP:
3535
		msr_info->data = svm->sysenter_esp;
A
Avi Kivity 已提交
3536
		break;
P
Paolo Bonzini 已提交
3537 3538 3539 3540 3541
	case MSR_TSC_AUX:
		if (!boot_cpu_has(X86_FEATURE_RDTSCP))
			return 1;
		msr_info->data = svm->tsc_aux;
		break;
J
Joerg Roedel 已提交
3542 3543 3544 3545 3546
	/*
	 * Nobody will change the following 5 values in the VMCB so we can
	 * safely return them on rdmsr. They will always be 0 until LBRV is
	 * implemented.
	 */
3547
	case MSR_IA32_DEBUGCTLMSR:
3548
		msr_info->data = svm->vmcb->save.dbgctl;
3549 3550
		break;
	case MSR_IA32_LASTBRANCHFROMIP:
3551
		msr_info->data = svm->vmcb->save.br_from;
3552 3553
		break;
	case MSR_IA32_LASTBRANCHTOIP:
3554
		msr_info->data = svm->vmcb->save.br_to;
3555 3556
		break;
	case MSR_IA32_LASTINTFROMIP:
3557
		msr_info->data = svm->vmcb->save.last_excp_from;
3558 3559
		break;
	case MSR_IA32_LASTINTTOIP:
3560
		msr_info->data = svm->vmcb->save.last_excp_to;
3561
		break;
A
Alexander Graf 已提交
3562
	case MSR_VM_HSAVE_PA:
3563
		msr_info->data = svm->nested.hsave_msr;
A
Alexander Graf 已提交
3564
		break;
3565
	case MSR_VM_CR:
3566
		msr_info->data = svm->nested.vm_cr_msr;
3567
		break;
3568
	case MSR_IA32_UCODE_REV:
3569
		msr_info->data = 0x01000065;
3570
		break;
3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587
	case MSR_F15H_IC_CFG: {

		int family, model;

		family = guest_cpuid_family(vcpu);
		model  = guest_cpuid_model(vcpu);

		if (family < 0 || model < 0)
			return kvm_get_msr_common(vcpu, msr_info);

		msr_info->data = 0;

		if (family == 0x15 &&
		    (model >= 0x2 && model < 0x20))
			msr_info->data = 0x1E;
		}
		break;
A
Avi Kivity 已提交
3588
	default:
3589
		return kvm_get_msr_common(vcpu, msr_info);
A
Avi Kivity 已提交
3590 3591 3592 3593
	}
	return 0;
}

A
Avi Kivity 已提交
3594
static int rdmsr_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3595
{
3596
	u32 ecx = kvm_register_read(&svm->vcpu, VCPU_REGS_RCX);
3597
	struct msr_data msr_info;
A
Avi Kivity 已提交
3598

3599 3600 3601
	msr_info.index = ecx;
	msr_info.host_initiated = false;
	if (svm_get_msr(&svm->vcpu, &msr_info)) {
3602
		trace_kvm_msr_read_ex(ecx);
3603
		kvm_inject_gp(&svm->vcpu, 0);
3604
		return 1;
3605
	} else {
3606
		trace_kvm_msr_read(ecx, msr_info.data);
3607

3608 3609 3610 3611
		kvm_register_write(&svm->vcpu, VCPU_REGS_RAX,
				   msr_info.data & 0xffffffff);
		kvm_register_write(&svm->vcpu, VCPU_REGS_RDX,
				   msr_info.data >> 32);
3612
		svm->next_rip = kvm_rip_read(&svm->vcpu) + 2;
3613
		return kvm_skip_emulated_instruction(&svm->vcpu);
A
Avi Kivity 已提交
3614 3615 3616
	}
}

3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641
static int svm_set_vm_cr(struct kvm_vcpu *vcpu, u64 data)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	int svm_dis, chg_mask;

	if (data & ~SVM_VM_CR_VALID_MASK)
		return 1;

	chg_mask = SVM_VM_CR_VALID_MASK;

	if (svm->nested.vm_cr_msr & SVM_VM_CR_SVM_DIS_MASK)
		chg_mask &= ~(SVM_VM_CR_SVM_LOCK_MASK | SVM_VM_CR_SVM_DIS_MASK);

	svm->nested.vm_cr_msr &= ~chg_mask;
	svm->nested.vm_cr_msr |= (data & chg_mask);

	svm_dis = svm->nested.vm_cr_msr & SVM_VM_CR_SVM_DIS_MASK;

	/* check for svm_disable while efer.svme is set */
	if (svm_dis && (vcpu->arch.efer & EFER_SVME))
		return 1;

	return 0;
}

3642
static int svm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr)
A
Avi Kivity 已提交
3643
{
3644 3645
	struct vcpu_svm *svm = to_svm(vcpu);

3646 3647
	u32 ecx = msr->index;
	u64 data = msr->data;
A
Avi Kivity 已提交
3648
	switch (ecx) {
3649
	case MSR_IA32_TSC:
3650
		kvm_write_tsc(vcpu, msr);
A
Avi Kivity 已提交
3651
		break;
B
Brian Gerst 已提交
3652
	case MSR_STAR:
3653
		svm->vmcb->save.star = data;
A
Avi Kivity 已提交
3654
		break;
3655
#ifdef CONFIG_X86_64
A
Avi Kivity 已提交
3656
	case MSR_LSTAR:
3657
		svm->vmcb->save.lstar = data;
A
Avi Kivity 已提交
3658 3659
		break;
	case MSR_CSTAR:
3660
		svm->vmcb->save.cstar = data;
A
Avi Kivity 已提交
3661 3662
		break;
	case MSR_KERNEL_GS_BASE:
3663
		svm->vmcb->save.kernel_gs_base = data;
A
Avi Kivity 已提交
3664 3665
		break;
	case MSR_SYSCALL_MASK:
3666
		svm->vmcb->save.sfmask = data;
A
Avi Kivity 已提交
3667 3668 3669
		break;
#endif
	case MSR_IA32_SYSENTER_CS:
3670
		svm->vmcb->save.sysenter_cs = data;
A
Avi Kivity 已提交
3671 3672
		break;
	case MSR_IA32_SYSENTER_EIP:
3673
		svm->sysenter_eip = data;
3674
		svm->vmcb->save.sysenter_eip = data;
A
Avi Kivity 已提交
3675 3676
		break;
	case MSR_IA32_SYSENTER_ESP:
3677
		svm->sysenter_esp = data;
3678
		svm->vmcb->save.sysenter_esp = data;
A
Avi Kivity 已提交
3679
		break;
P
Paolo Bonzini 已提交
3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691
	case MSR_TSC_AUX:
		if (!boot_cpu_has(X86_FEATURE_RDTSCP))
			return 1;

		/*
		 * This is rare, so we update the MSR here instead of using
		 * direct_access_msrs.  Doing that would require a rdmsr in
		 * svm_vcpu_put.
		 */
		svm->tsc_aux = data;
		wrmsrl(MSR_TSC_AUX, svm->tsc_aux);
		break;
3692
	case MSR_IA32_DEBUGCTLMSR:
3693
		if (!boot_cpu_has(X86_FEATURE_LBRV)) {
3694 3695
			vcpu_unimpl(vcpu, "%s: MSR_IA32_DEBUGCTL 0x%llx, nop\n",
				    __func__, data);
3696 3697 3698 3699 3700 3701
			break;
		}
		if (data & DEBUGCTL_RESERVED_BITS)
			return 1;

		svm->vmcb->save.dbgctl = data;
3702
		mark_dirty(svm->vmcb, VMCB_LBR);
3703 3704 3705 3706
		if (data & (1ULL<<0))
			svm_enable_lbrv(svm);
		else
			svm_disable_lbrv(svm);
3707
		break;
A
Alexander Graf 已提交
3708
	case MSR_VM_HSAVE_PA:
3709
		svm->nested.hsave_msr = data;
3710
		break;
3711
	case MSR_VM_CR:
3712
		return svm_set_vm_cr(vcpu, data);
3713
	case MSR_VM_IGNNE:
3714
		vcpu_unimpl(vcpu, "unimplemented wrmsr: 0x%x data 0x%llx\n", ecx, data);
3715
		break;
3716 3717 3718 3719
	case MSR_IA32_APICBASE:
		if (kvm_vcpu_apicv_active(vcpu))
			avic_update_vapic_bar(to_svm(vcpu), data);
		/* Follow through */
A
Avi Kivity 已提交
3720
	default:
3721
		return kvm_set_msr_common(vcpu, msr);
A
Avi Kivity 已提交
3722 3723 3724 3725
	}
	return 0;
}

A
Avi Kivity 已提交
3726
static int wrmsr_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3727
{
3728
	struct msr_data msr;
3729 3730
	u32 ecx = kvm_register_read(&svm->vcpu, VCPU_REGS_RCX);
	u64 data = kvm_read_edx_eax(&svm->vcpu);
3731

3732 3733 3734
	msr.data = data;
	msr.index = ecx;
	msr.host_initiated = false;
3735

3736
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 2;
3737
	if (kvm_set_msr(&svm->vcpu, &msr)) {
3738
		trace_kvm_msr_write_ex(ecx, data);
3739
		kvm_inject_gp(&svm->vcpu, 0);
3740
		return 1;
3741 3742
	} else {
		trace_kvm_msr_write(ecx, data);
3743
		return kvm_skip_emulated_instruction(&svm->vcpu);
3744
	}
A
Avi Kivity 已提交
3745 3746
}

A
Avi Kivity 已提交
3747
static int msr_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3748
{
R
Rusty Russell 已提交
3749
	if (svm->vmcb->control.exit_info_1)
A
Avi Kivity 已提交
3750
		return wrmsr_interception(svm);
A
Avi Kivity 已提交
3751
	else
A
Avi Kivity 已提交
3752
		return rdmsr_interception(svm);
A
Avi Kivity 已提交
3753 3754
}

A
Avi Kivity 已提交
3755
static int interrupt_window_interception(struct vcpu_svm *svm)
3756
{
3757
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
3758
	svm_clear_vintr(svm);
3759
	svm->vmcb->control.int_ctl &= ~V_IRQ_MASK;
3760
	mark_dirty(svm->vmcb, VMCB_INTR);
3761
	++svm->vcpu.stat.irq_window_exits;
3762 3763 3764
	return 1;
}

3765 3766
static int pause_interception(struct vcpu_svm *svm)
{
3767 3768 3769 3770
	struct kvm_vcpu *vcpu = &svm->vcpu;
	bool in_kernel = (svm_get_cpl(vcpu) == 0);

	kvm_vcpu_on_spin(vcpu, in_kernel);
3771 3772 3773
	return 1;
}

3774 3775
static int nop_interception(struct vcpu_svm *svm)
{
3776
	return kvm_skip_emulated_instruction(&(svm->vcpu));
3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790
}

static int monitor_interception(struct vcpu_svm *svm)
{
	printk_once(KERN_WARNING "kvm: MONITOR instruction emulated as NOP!\n");
	return nop_interception(svm);
}

static int mwait_interception(struct vcpu_svm *svm)
{
	printk_once(KERN_WARNING "kvm: MWAIT instruction emulated as NOP!\n");
	return nop_interception(svm);
}

3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802
enum avic_ipi_failure_cause {
	AVIC_IPI_FAILURE_INVALID_INT_TYPE,
	AVIC_IPI_FAILURE_TARGET_NOT_RUNNING,
	AVIC_IPI_FAILURE_INVALID_TARGET,
	AVIC_IPI_FAILURE_INVALID_BACKING_PAGE,
};

static int avic_incomplete_ipi_interception(struct vcpu_svm *svm)
{
	u32 icrh = svm->vmcb->control.exit_info_1 >> 32;
	u32 icrl = svm->vmcb->control.exit_info_1;
	u32 id = svm->vmcb->control.exit_info_2 >> 32;
D
Dan Carpenter 已提交
3803
	u32 index = svm->vmcb->control.exit_info_2 & 0xFF;
3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062
	struct kvm_lapic *apic = svm->vcpu.arch.apic;

	trace_kvm_avic_incomplete_ipi(svm->vcpu.vcpu_id, icrh, icrl, id, index);

	switch (id) {
	case AVIC_IPI_FAILURE_INVALID_INT_TYPE:
		/*
		 * AVIC hardware handles the generation of
		 * IPIs when the specified Message Type is Fixed
		 * (also known as fixed delivery mode) and
		 * the Trigger Mode is edge-triggered. The hardware
		 * also supports self and broadcast delivery modes
		 * specified via the Destination Shorthand(DSH)
		 * field of the ICRL. Logical and physical APIC ID
		 * formats are supported. All other IPI types cause
		 * a #VMEXIT, which needs to emulated.
		 */
		kvm_lapic_reg_write(apic, APIC_ICR2, icrh);
		kvm_lapic_reg_write(apic, APIC_ICR, icrl);
		break;
	case AVIC_IPI_FAILURE_TARGET_NOT_RUNNING: {
		int i;
		struct kvm_vcpu *vcpu;
		struct kvm *kvm = svm->vcpu.kvm;
		struct kvm_lapic *apic = svm->vcpu.arch.apic;

		/*
		 * At this point, we expect that the AVIC HW has already
		 * set the appropriate IRR bits on the valid target
		 * vcpus. So, we just need to kick the appropriate vcpu.
		 */
		kvm_for_each_vcpu(i, vcpu, kvm) {
			bool m = kvm_apic_match_dest(vcpu, apic,
						     icrl & KVM_APIC_SHORT_MASK,
						     GET_APIC_DEST_FIELD(icrh),
						     icrl & KVM_APIC_DEST_MASK);

			if (m && !avic_vcpu_is_running(vcpu))
				kvm_vcpu_wake_up(vcpu);
		}
		break;
	}
	case AVIC_IPI_FAILURE_INVALID_TARGET:
		break;
	case AVIC_IPI_FAILURE_INVALID_BACKING_PAGE:
		WARN_ONCE(1, "Invalid backing page\n");
		break;
	default:
		pr_err("Unknown IPI interception\n");
	}

	return 1;
}

static u32 *avic_get_logical_id_entry(struct kvm_vcpu *vcpu, u32 ldr, bool flat)
{
	struct kvm_arch *vm_data = &vcpu->kvm->arch;
	int index;
	u32 *logical_apic_id_table;
	int dlid = GET_APIC_LOGICAL_ID(ldr);

	if (!dlid)
		return NULL;

	if (flat) { /* flat */
		index = ffs(dlid) - 1;
		if (index > 7)
			return NULL;
	} else { /* cluster */
		int cluster = (dlid & 0xf0) >> 4;
		int apic = ffs(dlid & 0x0f) - 1;

		if ((apic < 0) || (apic > 7) ||
		    (cluster >= 0xf))
			return NULL;
		index = (cluster << 2) + apic;
	}

	logical_apic_id_table = (u32 *) page_address(vm_data->avic_logical_id_table_page);

	return &logical_apic_id_table[index];
}

static int avic_ldr_write(struct kvm_vcpu *vcpu, u8 g_physical_id, u32 ldr,
			  bool valid)
{
	bool flat;
	u32 *entry, new_entry;

	flat = kvm_lapic_get_reg(vcpu->arch.apic, APIC_DFR) == APIC_DFR_FLAT;
	entry = avic_get_logical_id_entry(vcpu, ldr, flat);
	if (!entry)
		return -EINVAL;

	new_entry = READ_ONCE(*entry);
	new_entry &= ~AVIC_LOGICAL_ID_ENTRY_GUEST_PHYSICAL_ID_MASK;
	new_entry |= (g_physical_id & AVIC_LOGICAL_ID_ENTRY_GUEST_PHYSICAL_ID_MASK);
	if (valid)
		new_entry |= AVIC_LOGICAL_ID_ENTRY_VALID_MASK;
	else
		new_entry &= ~AVIC_LOGICAL_ID_ENTRY_VALID_MASK;
	WRITE_ONCE(*entry, new_entry);

	return 0;
}

static int avic_handle_ldr_update(struct kvm_vcpu *vcpu)
{
	int ret;
	struct vcpu_svm *svm = to_svm(vcpu);
	u32 ldr = kvm_lapic_get_reg(vcpu->arch.apic, APIC_LDR);

	if (!ldr)
		return 1;

	ret = avic_ldr_write(vcpu, vcpu->vcpu_id, ldr, true);
	if (ret && svm->ldr_reg) {
		avic_ldr_write(vcpu, 0, svm->ldr_reg, false);
		svm->ldr_reg = 0;
	} else {
		svm->ldr_reg = ldr;
	}
	return ret;
}

static int avic_handle_apic_id_update(struct kvm_vcpu *vcpu)
{
	u64 *old, *new;
	struct vcpu_svm *svm = to_svm(vcpu);
	u32 apic_id_reg = kvm_lapic_get_reg(vcpu->arch.apic, APIC_ID);
	u32 id = (apic_id_reg >> 24) & 0xff;

	if (vcpu->vcpu_id == id)
		return 0;

	old = avic_get_physical_id_entry(vcpu, vcpu->vcpu_id);
	new = avic_get_physical_id_entry(vcpu, id);
	if (!new || !old)
		return 1;

	/* We need to move physical_id_entry to new offset */
	*new = *old;
	*old = 0ULL;
	to_svm(vcpu)->avic_physical_id_cache = new;

	/*
	 * Also update the guest physical APIC ID in the logical
	 * APIC ID table entry if already setup the LDR.
	 */
	if (svm->ldr_reg)
		avic_handle_ldr_update(vcpu);

	return 0;
}

static int avic_handle_dfr_update(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	struct kvm_arch *vm_data = &vcpu->kvm->arch;
	u32 dfr = kvm_lapic_get_reg(vcpu->arch.apic, APIC_DFR);
	u32 mod = (dfr >> 28) & 0xf;

	/*
	 * We assume that all local APICs are using the same type.
	 * If this changes, we need to flush the AVIC logical
	 * APID id table.
	 */
	if (vm_data->ldr_mode == mod)
		return 0;

	clear_page(page_address(vm_data->avic_logical_id_table_page));
	vm_data->ldr_mode = mod;

	if (svm->ldr_reg)
		avic_handle_ldr_update(vcpu);
	return 0;
}

static int avic_unaccel_trap_write(struct vcpu_svm *svm)
{
	struct kvm_lapic *apic = svm->vcpu.arch.apic;
	u32 offset = svm->vmcb->control.exit_info_1 &
				AVIC_UNACCEL_ACCESS_OFFSET_MASK;

	switch (offset) {
	case APIC_ID:
		if (avic_handle_apic_id_update(&svm->vcpu))
			return 0;
		break;
	case APIC_LDR:
		if (avic_handle_ldr_update(&svm->vcpu))
			return 0;
		break;
	case APIC_DFR:
		avic_handle_dfr_update(&svm->vcpu);
		break;
	default:
		break;
	}

	kvm_lapic_reg_write(apic, offset, kvm_lapic_get_reg(apic, offset));

	return 1;
}

static bool is_avic_unaccelerated_access_trap(u32 offset)
{
	bool ret = false;

	switch (offset) {
	case APIC_ID:
	case APIC_EOI:
	case APIC_RRR:
	case APIC_LDR:
	case APIC_DFR:
	case APIC_SPIV:
	case APIC_ESR:
	case APIC_ICR:
	case APIC_LVTT:
	case APIC_LVTTHMR:
	case APIC_LVTPC:
	case APIC_LVT0:
	case APIC_LVT1:
	case APIC_LVTERR:
	case APIC_TMICT:
	case APIC_TDCR:
		ret = true;
		break;
	default:
		break;
	}
	return ret;
}

static int avic_unaccelerated_access_interception(struct vcpu_svm *svm)
{
	int ret = 0;
	u32 offset = svm->vmcb->control.exit_info_1 &
		     AVIC_UNACCEL_ACCESS_OFFSET_MASK;
	u32 vector = svm->vmcb->control.exit_info_2 &
		     AVIC_UNACCEL_ACCESS_VECTOR_MASK;
	bool write = (svm->vmcb->control.exit_info_1 >> 32) &
		     AVIC_UNACCEL_ACCESS_WRITE_MASK;
	bool trap = is_avic_unaccelerated_access_trap(offset);

	trace_kvm_avic_unaccelerated_access(svm->vcpu.vcpu_id, offset,
					    trap, write, vector);
	if (trap) {
		/* Handling Trap */
		WARN_ONCE(!write, "svm: Handling trap read.\n");
		ret = avic_unaccel_trap_write(svm);
	} else {
		/* Handling Fault */
		ret = (emulate_instruction(&svm->vcpu, 0) == EMULATE_DONE);
	}

	return ret;
}

4063
static int (*const svm_exit_handlers[])(struct vcpu_svm *svm) = {
4064 4065 4066 4067
	[SVM_EXIT_READ_CR0]			= cr_interception,
	[SVM_EXIT_READ_CR3]			= cr_interception,
	[SVM_EXIT_READ_CR4]			= cr_interception,
	[SVM_EXIT_READ_CR8]			= cr_interception,
4068
	[SVM_EXIT_CR0_SEL_WRITE]		= cr_interception,
4069
	[SVM_EXIT_WRITE_CR0]			= cr_interception,
4070 4071
	[SVM_EXIT_WRITE_CR3]			= cr_interception,
	[SVM_EXIT_WRITE_CR4]			= cr_interception,
J
Joerg Roedel 已提交
4072
	[SVM_EXIT_WRITE_CR8]			= cr8_write_interception,
4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088
	[SVM_EXIT_READ_DR0]			= dr_interception,
	[SVM_EXIT_READ_DR1]			= dr_interception,
	[SVM_EXIT_READ_DR2]			= dr_interception,
	[SVM_EXIT_READ_DR3]			= dr_interception,
	[SVM_EXIT_READ_DR4]			= dr_interception,
	[SVM_EXIT_READ_DR5]			= dr_interception,
	[SVM_EXIT_READ_DR6]			= dr_interception,
	[SVM_EXIT_READ_DR7]			= dr_interception,
	[SVM_EXIT_WRITE_DR0]			= dr_interception,
	[SVM_EXIT_WRITE_DR1]			= dr_interception,
	[SVM_EXIT_WRITE_DR2]			= dr_interception,
	[SVM_EXIT_WRITE_DR3]			= dr_interception,
	[SVM_EXIT_WRITE_DR4]			= dr_interception,
	[SVM_EXIT_WRITE_DR5]			= dr_interception,
	[SVM_EXIT_WRITE_DR6]			= dr_interception,
	[SVM_EXIT_WRITE_DR7]			= dr_interception,
J
Jan Kiszka 已提交
4089 4090
	[SVM_EXIT_EXCP_BASE + DB_VECTOR]	= db_interception,
	[SVM_EXIT_EXCP_BASE + BP_VECTOR]	= bp_interception,
4091
	[SVM_EXIT_EXCP_BASE + UD_VECTOR]	= ud_interception,
J
Joerg Roedel 已提交
4092 4093
	[SVM_EXIT_EXCP_BASE + PF_VECTOR]	= pf_interception,
	[SVM_EXIT_EXCP_BASE + MC_VECTOR]	= mc_interception,
4094
	[SVM_EXIT_EXCP_BASE + AC_VECTOR]	= ac_interception,
J
Joerg Roedel 已提交
4095
	[SVM_EXIT_INTR]				= intr_interception,
4096
	[SVM_EXIT_NMI]				= nmi_interception,
A
Avi Kivity 已提交
4097 4098
	[SVM_EXIT_SMI]				= nop_on_interception,
	[SVM_EXIT_INIT]				= nop_on_interception,
4099
	[SVM_EXIT_VINTR]			= interrupt_window_interception,
A
Avi Kivity 已提交
4100
	[SVM_EXIT_RDPMC]			= rdpmc_interception,
A
Avi Kivity 已提交
4101
	[SVM_EXIT_CPUID]			= cpuid_interception,
4102
	[SVM_EXIT_IRET]                         = iret_interception,
4103
	[SVM_EXIT_INVD]                         = emulate_on_interception,
4104
	[SVM_EXIT_PAUSE]			= pause_interception,
A
Avi Kivity 已提交
4105
	[SVM_EXIT_HLT]				= halt_interception,
M
Marcelo Tosatti 已提交
4106
	[SVM_EXIT_INVLPG]			= invlpg_interception,
A
Alexander Graf 已提交
4107
	[SVM_EXIT_INVLPGA]			= invlpga_interception,
J
Joerg Roedel 已提交
4108
	[SVM_EXIT_IOIO]				= io_interception,
A
Avi Kivity 已提交
4109 4110
	[SVM_EXIT_MSR]				= msr_interception,
	[SVM_EXIT_TASK_SWITCH]			= task_switch_interception,
4111
	[SVM_EXIT_SHUTDOWN]			= shutdown_interception,
A
Alexander Graf 已提交
4112
	[SVM_EXIT_VMRUN]			= vmrun_interception,
4113
	[SVM_EXIT_VMMCALL]			= vmmcall_interception,
4114 4115
	[SVM_EXIT_VMLOAD]			= vmload_interception,
	[SVM_EXIT_VMSAVE]			= vmsave_interception,
4116 4117
	[SVM_EXIT_STGI]				= stgi_interception,
	[SVM_EXIT_CLGI]				= clgi_interception,
4118
	[SVM_EXIT_SKINIT]			= skinit_interception,
D
David Kaplan 已提交
4119
	[SVM_EXIT_WBINVD]                       = wbinvd_interception,
4120 4121
	[SVM_EXIT_MONITOR]			= monitor_interception,
	[SVM_EXIT_MWAIT]			= mwait_interception,
J
Joerg Roedel 已提交
4122
	[SVM_EXIT_XSETBV]			= xsetbv_interception,
4123
	[SVM_EXIT_NPF]				= pf_interception,
P
Paolo Bonzini 已提交
4124
	[SVM_EXIT_RSM]                          = emulate_on_interception,
4125 4126
	[SVM_EXIT_AVIC_INCOMPLETE_IPI]		= avic_incomplete_ipi_interception,
	[SVM_EXIT_AVIC_UNACCELERATED_ACCESS]	= avic_unaccelerated_access_interception,
A
Avi Kivity 已提交
4127 4128
};

4129
static void dump_vmcb(struct kvm_vcpu *vcpu)
4130 4131 4132 4133 4134 4135
{
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb_control_area *control = &svm->vmcb->control;
	struct vmcb_save_area *save = &svm->vmcb->save;

	pr_err("VMCB Control Area:\n");
4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157
	pr_err("%-20s%04x\n", "cr_read:", control->intercept_cr & 0xffff);
	pr_err("%-20s%04x\n", "cr_write:", control->intercept_cr >> 16);
	pr_err("%-20s%04x\n", "dr_read:", control->intercept_dr & 0xffff);
	pr_err("%-20s%04x\n", "dr_write:", control->intercept_dr >> 16);
	pr_err("%-20s%08x\n", "exceptions:", control->intercept_exceptions);
	pr_err("%-20s%016llx\n", "intercepts:", control->intercept);
	pr_err("%-20s%d\n", "pause filter count:", control->pause_filter_count);
	pr_err("%-20s%016llx\n", "iopm_base_pa:", control->iopm_base_pa);
	pr_err("%-20s%016llx\n", "msrpm_base_pa:", control->msrpm_base_pa);
	pr_err("%-20s%016llx\n", "tsc_offset:", control->tsc_offset);
	pr_err("%-20s%d\n", "asid:", control->asid);
	pr_err("%-20s%d\n", "tlb_ctl:", control->tlb_ctl);
	pr_err("%-20s%08x\n", "int_ctl:", control->int_ctl);
	pr_err("%-20s%08x\n", "int_vector:", control->int_vector);
	pr_err("%-20s%08x\n", "int_state:", control->int_state);
	pr_err("%-20s%08x\n", "exit_code:", control->exit_code);
	pr_err("%-20s%016llx\n", "exit_info1:", control->exit_info_1);
	pr_err("%-20s%016llx\n", "exit_info2:", control->exit_info_2);
	pr_err("%-20s%08x\n", "exit_int_info:", control->exit_int_info);
	pr_err("%-20s%08x\n", "exit_int_info_err:", control->exit_int_info_err);
	pr_err("%-20s%lld\n", "nested_ctl:", control->nested_ctl);
	pr_err("%-20s%016llx\n", "nested_cr3:", control->nested_cr3);
4158
	pr_err("%-20s%016llx\n", "avic_vapic_bar:", control->avic_vapic_bar);
4159 4160
	pr_err("%-20s%08x\n", "event_inj:", control->event_inj);
	pr_err("%-20s%08x\n", "event_inj_err:", control->event_inj_err);
4161
	pr_err("%-20s%lld\n", "virt_ext:", control->virt_ext);
4162
	pr_err("%-20s%016llx\n", "next_rip:", control->next_rip);
4163 4164 4165
	pr_err("%-20s%016llx\n", "avic_backing_page:", control->avic_backing_page);
	pr_err("%-20s%016llx\n", "avic_logical_id:", control->avic_logical_id);
	pr_err("%-20s%016llx\n", "avic_physical_id:", control->avic_physical_id);
4166
	pr_err("VMCB State Save Area:\n");
4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "es:",
	       save->es.selector, save->es.attrib,
	       save->es.limit, save->es.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "cs:",
	       save->cs.selector, save->cs.attrib,
	       save->cs.limit, save->cs.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "ss:",
	       save->ss.selector, save->ss.attrib,
	       save->ss.limit, save->ss.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "ds:",
	       save->ds.selector, save->ds.attrib,
	       save->ds.limit, save->ds.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "fs:",
	       save->fs.selector, save->fs.attrib,
	       save->fs.limit, save->fs.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "gs:",
	       save->gs.selector, save->gs.attrib,
	       save->gs.limit, save->gs.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "gdtr:",
	       save->gdtr.selector, save->gdtr.attrib,
	       save->gdtr.limit, save->gdtr.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "ldtr:",
	       save->ldtr.selector, save->ldtr.attrib,
	       save->ldtr.limit, save->ldtr.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "idtr:",
	       save->idtr.selector, save->idtr.attrib,
	       save->idtr.limit, save->idtr.base);
	pr_err("%-5s s: %04x a: %04x l: %08x b: %016llx\n",
	       "tr:",
	       save->tr.selector, save->tr.attrib,
	       save->tr.limit, save->tr.base);
4207 4208
	pr_err("cpl:            %d                efer:         %016llx\n",
		save->cpl, save->efer);
4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "cr0:", save->cr0, "cr2:", save->cr2);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "cr3:", save->cr3, "cr4:", save->cr4);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "dr6:", save->dr6, "dr7:", save->dr7);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "rip:", save->rip, "rflags:", save->rflags);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "rsp:", save->rsp, "rax:", save->rax);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "star:", save->star, "lstar:", save->lstar);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "cstar:", save->cstar, "sfmask:", save->sfmask);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "kernel_gs_base:", save->kernel_gs_base,
	       "sysenter_cs:", save->sysenter_cs);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "sysenter_esp:", save->sysenter_esp,
	       "sysenter_eip:", save->sysenter_eip);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "gpat:", save->g_pat, "dbgctl:", save->dbgctl);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "br_from:", save->br_from, "br_to:", save->br_to);
	pr_err("%-15s %016llx %-13s %016llx\n",
	       "excp_from:", save->last_excp_from,
	       "excp_to:", save->last_excp_to);
4236 4237
}

4238 4239 4240 4241 4242 4243 4244 4245
static void svm_get_exit_info(struct kvm_vcpu *vcpu, u64 *info1, u64 *info2)
{
	struct vmcb_control_area *control = &to_svm(vcpu)->vmcb->control;

	*info1 = control->exit_info_1;
	*info2 = control->exit_info_2;
}

A
Avi Kivity 已提交
4246
static int handle_exit(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
4247
{
4248
	struct vcpu_svm *svm = to_svm(vcpu);
A
Avi Kivity 已提交
4249
	struct kvm_run *kvm_run = vcpu->run;
4250
	u32 exit_code = svm->vmcb->control.exit_code;
A
Avi Kivity 已提交
4251

4252 4253
	trace_kvm_exit(exit_code, vcpu, KVM_ISA_SVM);

4254
	if (!is_cr_intercept(svm, INTERCEPT_CR0_WRITE))
4255 4256 4257
		vcpu->arch.cr0 = svm->vmcb->save.cr0;
	if (npt_enabled)
		vcpu->arch.cr3 = svm->vmcb->save.cr3;
4258

4259 4260 4261 4262 4263 4264 4265
	if (unlikely(svm->nested.exit_required)) {
		nested_svm_vmexit(svm);
		svm->nested.exit_required = false;

		return 1;
	}

4266
	if (is_guest_mode(vcpu)) {
4267 4268
		int vmexit;

4269 4270 4271 4272
		trace_kvm_nested_vmexit(svm->vmcb->save.rip, exit_code,
					svm->vmcb->control.exit_info_1,
					svm->vmcb->control.exit_info_2,
					svm->vmcb->control.exit_int_info,
4273 4274
					svm->vmcb->control.exit_int_info_err,
					KVM_ISA_SVM);
4275

4276 4277 4278 4279 4280 4281
		vmexit = nested_svm_exit_special(svm);

		if (vmexit == NESTED_EXIT_CONTINUE)
			vmexit = nested_svm_exit_handled(svm);

		if (vmexit == NESTED_EXIT_DONE)
4282 4283 4284
			return 1;
	}

4285 4286
	svm_complete_interrupts(svm);

4287 4288 4289 4290
	if (svm->vmcb->control.exit_code == SVM_EXIT_ERR) {
		kvm_run->exit_reason = KVM_EXIT_FAIL_ENTRY;
		kvm_run->fail_entry.hardware_entry_failure_reason
			= svm->vmcb->control.exit_code;
4291 4292
		pr_err("KVM: FAILED VMRUN WITH VMCB:\n");
		dump_vmcb(vcpu);
4293 4294 4295
		return 0;
	}

4296
	if (is_external_interrupt(svm->vmcb->control.exit_int_info) &&
4297
	    exit_code != SVM_EXIT_EXCP_BASE + PF_VECTOR &&
4298 4299
	    exit_code != SVM_EXIT_NPF && exit_code != SVM_EXIT_TASK_SWITCH &&
	    exit_code != SVM_EXIT_INTR && exit_code != SVM_EXIT_NMI)
4300
		printk(KERN_ERR "%s: unexpected exit_int_info 0x%x "
A
Avi Kivity 已提交
4301
		       "exit_code 0x%x\n",
4302
		       __func__, svm->vmcb->control.exit_int_info,
A
Avi Kivity 已提交
4303 4304
		       exit_code);

4305
	if (exit_code >= ARRAY_SIZE(svm_exit_handlers)
J
Joe Perches 已提交
4306
	    || !svm_exit_handlers[exit_code]) {
4307
		WARN_ONCE(1, "svm: unexpected exit reason 0x%x\n", exit_code);
4308 4309
		kvm_queue_exception(vcpu, UD_VECTOR);
		return 1;
A
Avi Kivity 已提交
4310 4311
	}

A
Avi Kivity 已提交
4312
	return svm_exit_handlers[exit_code](svm);
A
Avi Kivity 已提交
4313 4314 4315 4316 4317 4318
}

static void reload_tss(struct kvm_vcpu *vcpu)
{
	int cpu = raw_smp_processor_id();

4319 4320
	struct svm_cpu_data *sd = per_cpu(svm_data, cpu);
	sd->tss_desc->type = 9; /* available 32/64-bit TSS */
A
Avi Kivity 已提交
4321 4322 4323
	load_TR_desc();
}

R
Rusty Russell 已提交
4324
static void pre_svm_run(struct vcpu_svm *svm)
A
Avi Kivity 已提交
4325 4326 4327
{
	int cpu = raw_smp_processor_id();

4328
	struct svm_cpu_data *sd = per_cpu(svm_data, cpu);
A
Avi Kivity 已提交
4329

4330
	/* FIXME: handle wraparound of asid_generation */
4331 4332
	if (svm->asid_generation != sd->asid_generation)
		new_asid(svm, sd);
A
Avi Kivity 已提交
4333 4334
}

4335 4336 4337 4338 4339 4340
static void svm_inject_nmi(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	svm->vmcb->control.event_inj = SVM_EVTINJ_VALID | SVM_EVTINJ_TYPE_NMI;
	vcpu->arch.hflags |= HF_NMI_MASK;
4341
	set_intercept(svm, INTERCEPT_IRET);
4342 4343
	++vcpu->stat.nmi_injections;
}
A
Avi Kivity 已提交
4344

4345
static inline void svm_inject_irq(struct vcpu_svm *svm, int irq)
A
Avi Kivity 已提交
4346 4347 4348
{
	struct vmcb_control_area *control;

4349
	/* The following fields are ignored when AVIC is enabled */
R
Rusty Russell 已提交
4350
	control = &svm->vmcb->control;
4351
	control->int_vector = irq;
A
Avi Kivity 已提交
4352 4353 4354
	control->int_ctl &= ~V_INTR_PRIO_MASK;
	control->int_ctl |= V_IRQ_MASK |
		((/*control->int_vector >> 4*/ 0xf) << V_INTR_PRIO_SHIFT);
4355
	mark_dirty(svm->vmcb, VMCB_INTR);
A
Avi Kivity 已提交
4356 4357
}

4358
static void svm_set_irq(struct kvm_vcpu *vcpu)
E
Eddie Dong 已提交
4359 4360 4361
{
	struct vcpu_svm *svm = to_svm(vcpu);

4362
	BUG_ON(!(gif_set(svm)));
4363

4364 4365 4366
	trace_kvm_inj_virq(vcpu->arch.interrupt.nr);
	++vcpu->stat.irq_injections;

4367 4368
	svm->vmcb->control.event_inj = vcpu->arch.interrupt.nr |
		SVM_EVTINJ_VALID | SVM_EVTINJ_TYPE_INTR;
E
Eddie Dong 已提交
4369 4370
}

4371 4372 4373 4374 4375
static inline bool svm_nested_virtualize_tpr(struct kvm_vcpu *vcpu)
{
	return is_guest_mode(vcpu) && (vcpu->arch.hflags & HF_VINTR_MASK);
}

4376
static void update_cr8_intercept(struct kvm_vcpu *vcpu, int tpr, int irr)
4377 4378 4379
{
	struct vcpu_svm *svm = to_svm(vcpu);

4380 4381
	if (svm_nested_virtualize_tpr(vcpu) ||
	    kvm_vcpu_apicv_active(vcpu))
4382 4383
		return;

4384 4385
	clr_cr_intercept(svm, INTERCEPT_CR8_WRITE);

4386
	if (irr == -1)
4387 4388
		return;

4389
	if (tpr >= irr)
4390
		set_cr_intercept(svm, INTERCEPT_CR8_WRITE);
4391
}
4392

4393 4394 4395 4396 4397
static void svm_set_virtual_x2apic_mode(struct kvm_vcpu *vcpu, bool set)
{
	return;
}

4398 4399
static bool svm_get_enable_apicv(void)
{
4400 4401 4402 4403 4404
	return avic;
}

static void svm_hwapic_irr_update(struct kvm_vcpu *vcpu, int max_irr)
{
4405 4406
}

4407
static void svm_hwapic_isr_update(struct kvm_vcpu *vcpu, int max_isr)
4408
{
4409 4410
}

4411
/* Note: Currently only used by Hyper-V. */
4412
static void svm_refresh_apicv_exec_ctrl(struct kvm_vcpu *vcpu)
4413
{
4414 4415 4416 4417 4418 4419 4420 4421
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb *vmcb = svm->vmcb;

	if (!avic)
		return;

	vmcb->control.int_ctl &= ~AVIC_ENABLE_MASK;
	mark_dirty(vmcb, VMCB_INTR);
4422 4423
}

4424
static void svm_load_eoi_exitmap(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
4425 4426 4427 4428
{
	return;
}

4429 4430 4431 4432 4433 4434 4435
static void svm_deliver_avic_intr(struct kvm_vcpu *vcpu, int vec)
{
	kvm_lapic_set_irr(vec, vcpu->arch.apic);
	smp_mb__after_atomic();

	if (avic_vcpu_is_running(vcpu))
		wrmsrl(SVM_AVIC_DOORBELL,
4436
		       kvm_cpu_get_apicid(vcpu->cpu));
4437 4438 4439 4440
	else
		kvm_vcpu_wake_up(vcpu);
}

4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529
static void svm_ir_list_del(struct vcpu_svm *svm, struct amd_iommu_pi_data *pi)
{
	unsigned long flags;
	struct amd_svm_iommu_ir *cur;

	spin_lock_irqsave(&svm->ir_list_lock, flags);
	list_for_each_entry(cur, &svm->ir_list, node) {
		if (cur->data != pi->ir_data)
			continue;
		list_del(&cur->node);
		kfree(cur);
		break;
	}
	spin_unlock_irqrestore(&svm->ir_list_lock, flags);
}

static int svm_ir_list_add(struct vcpu_svm *svm, struct amd_iommu_pi_data *pi)
{
	int ret = 0;
	unsigned long flags;
	struct amd_svm_iommu_ir *ir;

	/**
	 * In some cases, the existing irte is updaed and re-set,
	 * so we need to check here if it's already been * added
	 * to the ir_list.
	 */
	if (pi->ir_data && (pi->prev_ga_tag != 0)) {
		struct kvm *kvm = svm->vcpu.kvm;
		u32 vcpu_id = AVIC_GATAG_TO_VCPUID(pi->prev_ga_tag);
		struct kvm_vcpu *prev_vcpu = kvm_get_vcpu_by_id(kvm, vcpu_id);
		struct vcpu_svm *prev_svm;

		if (!prev_vcpu) {
			ret = -EINVAL;
			goto out;
		}

		prev_svm = to_svm(prev_vcpu);
		svm_ir_list_del(prev_svm, pi);
	}

	/**
	 * Allocating new amd_iommu_pi_data, which will get
	 * add to the per-vcpu ir_list.
	 */
	ir = kzalloc(sizeof(struct amd_svm_iommu_ir), GFP_KERNEL);
	if (!ir) {
		ret = -ENOMEM;
		goto out;
	}
	ir->data = pi->ir_data;

	spin_lock_irqsave(&svm->ir_list_lock, flags);
	list_add(&ir->node, &svm->ir_list);
	spin_unlock_irqrestore(&svm->ir_list_lock, flags);
out:
	return ret;
}

/**
 * Note:
 * The HW cannot support posting multicast/broadcast
 * interrupts to a vCPU. So, we still use legacy interrupt
 * remapping for these kind of interrupts.
 *
 * For lowest-priority interrupts, we only support
 * those with single CPU as the destination, e.g. user
 * configures the interrupts via /proc/irq or uses
 * irqbalance to make the interrupts single-CPU.
 */
static int
get_pi_vcpu_info(struct kvm *kvm, struct kvm_kernel_irq_routing_entry *e,
		 struct vcpu_data *vcpu_info, struct vcpu_svm **svm)
{
	struct kvm_lapic_irq irq;
	struct kvm_vcpu *vcpu = NULL;

	kvm_set_msi_irq(kvm, e, &irq);

	if (!kvm_intr_is_single_vcpu(kvm, &irq, &vcpu)) {
		pr_debug("SVM: %s: use legacy intr remap mode for irq %u\n",
			 __func__, irq.vector);
		return -1;
	}

	pr_debug("SVM: %s: use GA mode for irq %u\n", __func__,
		 irq.vector);
	*svm = to_svm(vcpu);
4530
	vcpu_info->pi_desc_addr = __sme_set(page_to_phys((*svm)->avic_backing_page));
4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580
	vcpu_info->vector = irq.vector;

	return 0;
}

/*
 * svm_update_pi_irte - set IRTE for Posted-Interrupts
 *
 * @kvm: kvm
 * @host_irq: host irq of the interrupt
 * @guest_irq: gsi of the interrupt
 * @set: set or unset PI
 * returns 0 on success, < 0 on failure
 */
static int svm_update_pi_irte(struct kvm *kvm, unsigned int host_irq,
			      uint32_t guest_irq, bool set)
{
	struct kvm_kernel_irq_routing_entry *e;
	struct kvm_irq_routing_table *irq_rt;
	int idx, ret = -EINVAL;

	if (!kvm_arch_has_assigned_device(kvm) ||
	    !irq_remapping_cap(IRQ_POSTING_CAP))
		return 0;

	pr_debug("SVM: %s: host_irq=%#x, guest_irq=%#x, set=%#x\n",
		 __func__, host_irq, guest_irq, set);

	idx = srcu_read_lock(&kvm->irq_srcu);
	irq_rt = srcu_dereference(kvm->irq_routing, &kvm->irq_srcu);
	WARN_ON(guest_irq >= irq_rt->nr_rt_entries);

	hlist_for_each_entry(e, &irq_rt->map[guest_irq], link) {
		struct vcpu_data vcpu_info;
		struct vcpu_svm *svm = NULL;

		if (e->type != KVM_IRQ_ROUTING_MSI)
			continue;

		/**
		 * Here, we setup with legacy mode in the following cases:
		 * 1. When cannot target interrupt to a specific vcpu.
		 * 2. Unsetting posted interrupt.
		 * 3. APIC virtialization is disabled for the vcpu.
		 */
		if (!get_pi_vcpu_info(kvm, e, &vcpu_info, &svm) && set &&
		    kvm_vcpu_apicv_active(&svm->vcpu)) {
			struct amd_iommu_pi_data pi;

			/* Try to enable guest_mode in IRTE */
4581 4582
			pi.base = __sme_set(page_to_phys(svm->avic_backing_page) &
					    AVIC_HPA_MASK);
4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641 4642 4643 4644
			pi.ga_tag = AVIC_GATAG(kvm->arch.avic_vm_id,
						     svm->vcpu.vcpu_id);
			pi.is_guest_mode = true;
			pi.vcpu_data = &vcpu_info;
			ret = irq_set_vcpu_affinity(host_irq, &pi);

			/**
			 * Here, we successfully setting up vcpu affinity in
			 * IOMMU guest mode. Now, we need to store the posted
			 * interrupt information in a per-vcpu ir_list so that
			 * we can reference to them directly when we update vcpu
			 * scheduling information in IOMMU irte.
			 */
			if (!ret && pi.is_guest_mode)
				svm_ir_list_add(svm, &pi);
		} else {
			/* Use legacy mode in IRTE */
			struct amd_iommu_pi_data pi;

			/**
			 * Here, pi is used to:
			 * - Tell IOMMU to use legacy mode for this interrupt.
			 * - Retrieve ga_tag of prior interrupt remapping data.
			 */
			pi.is_guest_mode = false;
			ret = irq_set_vcpu_affinity(host_irq, &pi);

			/**
			 * Check if the posted interrupt was previously
			 * setup with the guest_mode by checking if the ga_tag
			 * was cached. If so, we need to clean up the per-vcpu
			 * ir_list.
			 */
			if (!ret && pi.prev_ga_tag) {
				int id = AVIC_GATAG_TO_VCPUID(pi.prev_ga_tag);
				struct kvm_vcpu *vcpu;

				vcpu = kvm_get_vcpu_by_id(kvm, id);
				if (vcpu)
					svm_ir_list_del(to_svm(vcpu), &pi);
			}
		}

		if (!ret && svm) {
			trace_kvm_pi_irte_update(svm->vcpu.vcpu_id,
						 host_irq, e->gsi,
						 vcpu_info.vector,
						 vcpu_info.pi_desc_addr, set);
		}

		if (ret < 0) {
			pr_err("%s: failed to update PI IRTE\n", __func__);
			goto out;
		}
	}

	ret = 0;
out:
	srcu_read_unlock(&kvm->irq_srcu, idx);
	return ret;
}

4645 4646 4647 4648
static int svm_nmi_allowed(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb *vmcb = svm->vmcb;
J
Joerg Roedel 已提交
4649 4650 4651 4652 4653 4654
	int ret;
	ret = !(vmcb->control.int_state & SVM_INTERRUPT_SHADOW_MASK) &&
	      !(svm->vcpu.arch.hflags & HF_NMI_MASK);
	ret = ret && gif_set(svm) && nested_svm_nmi(svm);

	return ret;
4655 4656
}

J
Jan Kiszka 已提交
4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669
static bool svm_get_nmi_mask(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	return !!(svm->vcpu.arch.hflags & HF_NMI_MASK);
}

static void svm_set_nmi_mask(struct kvm_vcpu *vcpu, bool masked)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	if (masked) {
		svm->vcpu.arch.hflags |= HF_NMI_MASK;
4670
		set_intercept(svm, INTERCEPT_IRET);
J
Jan Kiszka 已提交
4671 4672
	} else {
		svm->vcpu.arch.hflags &= ~HF_NMI_MASK;
4673
		clr_intercept(svm, INTERCEPT_IRET);
J
Jan Kiszka 已提交
4674 4675 4676
	}
}

4677 4678 4679 4680
static int svm_interrupt_allowed(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb *vmcb = svm->vmcb;
4681 4682 4683 4684 4685 4686
	int ret;

	if (!gif_set(svm) ||
	     (vmcb->control.int_state & SVM_INTERRUPT_SHADOW_MASK))
		return 0;

4687
	ret = !!(kvm_get_rflags(vcpu) & X86_EFLAGS_IF);
4688

4689
	if (is_guest_mode(vcpu))
4690 4691 4692
		return ret && !(svm->vcpu.arch.hflags & HF_VINTR_MASK);

	return ret;
4693 4694
}

4695
static void enable_irq_window(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
4696
{
4697 4698
	struct vcpu_svm *svm = to_svm(vcpu);

4699 4700 4701
	if (kvm_vcpu_apicv_active(vcpu))
		return;

J
Joerg Roedel 已提交
4702 4703 4704 4705
	/*
	 * In case GIF=0 we can't rely on the CPU to tell us when GIF becomes
	 * 1, because that's a separate STGI/VMRUN intercept.  The next time we
	 * get that intercept, this function will be called again though and
4706 4707 4708
	 * we'll get the vintr intercept. However, if the vGIF feature is
	 * enabled, the STGI interception will not occur. Enable the irq
	 * window under the assumption that the hardware will set the GIF.
J
Joerg Roedel 已提交
4709
	 */
4710
	if ((vgif_enabled(svm) || gif_set(svm)) && nested_svm_intr(svm)) {
4711 4712 4713
		svm_set_vintr(svm);
		svm_inject_irq(svm, 0x0);
	}
4714 4715
}

4716
static void enable_nmi_window(struct kvm_vcpu *vcpu)
4717
{
4718
	struct vcpu_svm *svm = to_svm(vcpu);
4719

4720 4721
	if ((svm->vcpu.arch.hflags & (HF_NMI_MASK | HF_IRET_MASK))
	    == HF_NMI_MASK)
4722
		return; /* IRET will cause a vm exit */
4723

4724 4725 4726
	if (!gif_set(svm)) {
		if (vgif_enabled(svm))
			set_intercept(svm, INTERCEPT_STGI);
4727
		return; /* STGI will cause a vm exit */
4728
	}
4729 4730 4731 4732

	if (svm->nested.exit_required)
		return; /* we're not going to run the guest yet */

J
Joerg Roedel 已提交
4733 4734 4735 4736
	/*
	 * Something prevents NMI from been injected. Single step over possible
	 * problem (IRET or exception injection or interrupt shadow)
	 */
4737
	svm->nmi_singlestep_guest_rflags = svm_get_rflags(vcpu);
J
Jan Kiszka 已提交
4738
	svm->nmi_singlestep = true;
4739
	svm->vmcb->save.rflags |= (X86_EFLAGS_TF | X86_EFLAGS_RF);
4740 4741
}

4742 4743 4744 4745 4746
static int svm_set_tss_addr(struct kvm *kvm, unsigned int addr)
{
	return 0;
}

4747 4748
static void svm_flush_tlb(struct kvm_vcpu *vcpu)
{
4749 4750 4751 4752 4753 4754
	struct vcpu_svm *svm = to_svm(vcpu);

	if (static_cpu_has(X86_FEATURE_FLUSHBYASID))
		svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ASID;
	else
		svm->asid_generation--;
4755 4756
}

4757 4758 4759 4760
static void svm_prepare_guest_switch(struct kvm_vcpu *vcpu)
{
}

4761 4762 4763 4764
static inline void sync_cr8_to_lapic(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

4765
	if (svm_nested_virtualize_tpr(vcpu))
4766 4767
		return;

4768
	if (!is_cr_intercept(svm, INTERCEPT_CR8_WRITE)) {
4769
		int cr8 = svm->vmcb->control.int_ctl & V_TPR_MASK;
4770
		kvm_set_cr8(vcpu, cr8);
4771 4772 4773
	}
}

4774 4775 4776 4777 4778
static inline void sync_lapic_to_cr8(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u64 cr8;

4779 4780
	if (svm_nested_virtualize_tpr(vcpu) ||
	    kvm_vcpu_apicv_active(vcpu))
4781 4782
		return;

4783 4784 4785 4786 4787
	cr8 = kvm_get_cr8(vcpu);
	svm->vmcb->control.int_ctl &= ~V_TPR_MASK;
	svm->vmcb->control.int_ctl |= cr8 & V_TPR_MASK;
}

4788 4789 4790 4791 4792
static void svm_complete_interrupts(struct vcpu_svm *svm)
{
	u8 vector;
	int type;
	u32 exitintinfo = svm->vmcb->control.exit_int_info;
4793 4794 4795
	unsigned int3_injected = svm->int3_injected;

	svm->int3_injected = 0;
4796

4797 4798 4799 4800 4801 4802
	/*
	 * If we've made progress since setting HF_IRET_MASK, we've
	 * executed an IRET and can allow NMI injection.
	 */
	if ((svm->vcpu.arch.hflags & HF_IRET_MASK)
	    && kvm_rip_read(&svm->vcpu) != svm->nmi_iret_rip) {
4803
		svm->vcpu.arch.hflags &= ~(HF_NMI_MASK | HF_IRET_MASK);
4804 4805
		kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
	}
4806

4807 4808 4809 4810 4811 4812 4813
	svm->vcpu.arch.nmi_injected = false;
	kvm_clear_exception_queue(&svm->vcpu);
	kvm_clear_interrupt_queue(&svm->vcpu);

	if (!(exitintinfo & SVM_EXITINTINFO_VALID))
		return;

4814 4815
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);

4816 4817 4818 4819 4820 4821 4822 4823
	vector = exitintinfo & SVM_EXITINTINFO_VEC_MASK;
	type = exitintinfo & SVM_EXITINTINFO_TYPE_MASK;

	switch (type) {
	case SVM_EXITINTINFO_TYPE_NMI:
		svm->vcpu.arch.nmi_injected = true;
		break;
	case SVM_EXITINTINFO_TYPE_EXEPT:
4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834
		/*
		 * In case of software exceptions, do not reinject the vector,
		 * but re-execute the instruction instead. Rewind RIP first
		 * if we emulated INT3 before.
		 */
		if (kvm_exception_is_soft(vector)) {
			if (vector == BP_VECTOR && int3_injected &&
			    kvm_is_linear_rip(&svm->vcpu, svm->int3_rip))
				kvm_rip_write(&svm->vcpu,
					      kvm_rip_read(&svm->vcpu) -
					      int3_injected);
4835
			break;
4836
		}
4837 4838
		if (exitintinfo & SVM_EXITINTINFO_VALID_ERR) {
			u32 err = svm->vmcb->control.exit_int_info_err;
4839
			kvm_requeue_exception_e(&svm->vcpu, vector, err);
4840 4841

		} else
4842
			kvm_requeue_exception(&svm->vcpu, vector);
4843 4844
		break;
	case SVM_EXITINTINFO_TYPE_INTR:
4845
		kvm_queue_interrupt(&svm->vcpu, vector, false);
4846 4847 4848 4849 4850 4851
		break;
	default:
		break;
	}
}

A
Avi Kivity 已提交
4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862
static void svm_cancel_injection(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb_control_area *control = &svm->vmcb->control;

	control->exit_int_info = control->event_inj;
	control->exit_int_info_err = control->event_inj_err;
	control->event_inj = 0;
	svm_complete_interrupts(svm);
}

A
Avi Kivity 已提交
4863
static void svm_vcpu_run(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
4864
{
4865
	struct vcpu_svm *svm = to_svm(vcpu);
4866

4867 4868 4869 4870
	svm->vmcb->save.rax = vcpu->arch.regs[VCPU_REGS_RAX];
	svm->vmcb->save.rsp = vcpu->arch.regs[VCPU_REGS_RSP];
	svm->vmcb->save.rip = vcpu->arch.regs[VCPU_REGS_RIP];

4871 4872 4873 4874 4875 4876 4877
	/*
	 * A vmexit emulation is required before the vcpu can be executed
	 * again.
	 */
	if (unlikely(svm->nested.exit_required))
		return;

4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888 4889 4890 4891 4892 4893
	/*
	 * Disable singlestep if we're injecting an interrupt/exception.
	 * We don't want our modified rflags to be pushed on the stack where
	 * we might not be able to easily reset them if we disabled NMI
	 * singlestep later.
	 */
	if (svm->nmi_singlestep && svm->vmcb->control.event_inj) {
		/*
		 * Event injection happens before external interrupts cause a
		 * vmexit and interrupts are disabled here, so smp_send_reschedule
		 * is enough to force an immediate vmexit.
		 */
		disable_nmi_singlestep(svm);
		smp_send_reschedule(vcpu->cpu);
	}

R
Rusty Russell 已提交
4894
	pre_svm_run(svm);
A
Avi Kivity 已提交
4895

4896 4897
	sync_lapic_to_cr8(vcpu);

4898
	svm->vmcb->save.cr2 = vcpu->arch.cr2;
A
Avi Kivity 已提交
4899

4900 4901 4902
	clgi();

	local_irq_enable();
4903

A
Avi Kivity 已提交
4904
	asm volatile (
A
Avi Kivity 已提交
4905 4906 4907 4908 4909 4910 4911
		"push %%" _ASM_BP "; \n\t"
		"mov %c[rbx](%[svm]), %%" _ASM_BX " \n\t"
		"mov %c[rcx](%[svm]), %%" _ASM_CX " \n\t"
		"mov %c[rdx](%[svm]), %%" _ASM_DX " \n\t"
		"mov %c[rsi](%[svm]), %%" _ASM_SI " \n\t"
		"mov %c[rdi](%[svm]), %%" _ASM_DI " \n\t"
		"mov %c[rbp](%[svm]), %%" _ASM_BP " \n\t"
4912
#ifdef CONFIG_X86_64
R
Rusty Russell 已提交
4913 4914 4915 4916 4917 4918 4919 4920
		"mov %c[r8](%[svm]),  %%r8  \n\t"
		"mov %c[r9](%[svm]),  %%r9  \n\t"
		"mov %c[r10](%[svm]), %%r10 \n\t"
		"mov %c[r11](%[svm]), %%r11 \n\t"
		"mov %c[r12](%[svm]), %%r12 \n\t"
		"mov %c[r13](%[svm]), %%r13 \n\t"
		"mov %c[r14](%[svm]), %%r14 \n\t"
		"mov %c[r15](%[svm]), %%r15 \n\t"
A
Avi Kivity 已提交
4921 4922 4923
#endif

		/* Enter guest mode */
A
Avi Kivity 已提交
4924 4925
		"push %%" _ASM_AX " \n\t"
		"mov %c[vmcb](%[svm]), %%" _ASM_AX " \n\t"
4926 4927 4928
		__ex(SVM_VMLOAD) "\n\t"
		__ex(SVM_VMRUN) "\n\t"
		__ex(SVM_VMSAVE) "\n\t"
A
Avi Kivity 已提交
4929
		"pop %%" _ASM_AX " \n\t"
A
Avi Kivity 已提交
4930 4931

		/* Save guest registers, load host registers */
A
Avi Kivity 已提交
4932 4933 4934 4935 4936 4937
		"mov %%" _ASM_BX ", %c[rbx](%[svm]) \n\t"
		"mov %%" _ASM_CX ", %c[rcx](%[svm]) \n\t"
		"mov %%" _ASM_DX ", %c[rdx](%[svm]) \n\t"
		"mov %%" _ASM_SI ", %c[rsi](%[svm]) \n\t"
		"mov %%" _ASM_DI ", %c[rdi](%[svm]) \n\t"
		"mov %%" _ASM_BP ", %c[rbp](%[svm]) \n\t"
4938
#ifdef CONFIG_X86_64
R
Rusty Russell 已提交
4939 4940 4941 4942 4943 4944 4945 4946
		"mov %%r8,  %c[r8](%[svm]) \n\t"
		"mov %%r9,  %c[r9](%[svm]) \n\t"
		"mov %%r10, %c[r10](%[svm]) \n\t"
		"mov %%r11, %c[r11](%[svm]) \n\t"
		"mov %%r12, %c[r12](%[svm]) \n\t"
		"mov %%r13, %c[r13](%[svm]) \n\t"
		"mov %%r14, %c[r14](%[svm]) \n\t"
		"mov %%r15, %c[r15](%[svm]) \n\t"
A
Avi Kivity 已提交
4947
#endif
A
Avi Kivity 已提交
4948
		"pop %%" _ASM_BP
A
Avi Kivity 已提交
4949
		:
R
Rusty Russell 已提交
4950
		: [svm]"a"(svm),
A
Avi Kivity 已提交
4951
		  [vmcb]"i"(offsetof(struct vcpu_svm, vmcb_pa)),
4952 4953 4954 4955 4956 4957
		  [rbx]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_RBX])),
		  [rcx]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_RCX])),
		  [rdx]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_RDX])),
		  [rsi]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_RSI])),
		  [rdi]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_RDI])),
		  [rbp]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_RBP]))
4958
#ifdef CONFIG_X86_64
4959 4960 4961 4962 4963 4964 4965 4966
		  , [r8]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R8])),
		  [r9]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R9])),
		  [r10]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R10])),
		  [r11]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R11])),
		  [r12]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R12])),
		  [r13]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R13])),
		  [r14]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R14])),
		  [r15]"i"(offsetof(struct vcpu_svm, vcpu.arch.regs[VCPU_REGS_R15]))
A
Avi Kivity 已提交
4967
#endif
4968 4969
		: "cc", "memory"
#ifdef CONFIG_X86_64
A
Avi Kivity 已提交
4970
		, "rbx", "rcx", "rdx", "rsi", "rdi"
4971
		, "r8", "r9", "r10", "r11" , "r12", "r13", "r14", "r15"
A
Avi Kivity 已提交
4972 4973
#else
		, "ebx", "ecx", "edx", "esi", "edi"
4974 4975
#endif
		);
A
Avi Kivity 已提交
4976

4977 4978 4979
#ifdef CONFIG_X86_64
	wrmsrl(MSR_GS_BASE, svm->host.gs_base);
#else
4980
	loadsegment(fs, svm->host.fs);
4981 4982 4983
#ifndef CONFIG_X86_32_LAZY_GS
	loadsegment(gs, svm->host.gs);
#endif
4984
#endif
A
Avi Kivity 已提交
4985 4986 4987

	reload_tss(vcpu);

4988 4989
	local_irq_disable();

4990 4991 4992 4993 4994
	vcpu->arch.cr2 = svm->vmcb->save.cr2;
	vcpu->arch.regs[VCPU_REGS_RAX] = svm->vmcb->save.rax;
	vcpu->arch.regs[VCPU_REGS_RSP] = svm->vmcb->save.rsp;
	vcpu->arch.regs[VCPU_REGS_RIP] = svm->vmcb->save.rip;

4995 4996 4997 4998 4999 5000 5001 5002 5003 5004
	if (unlikely(svm->vmcb->control.exit_code == SVM_EXIT_NMI))
		kvm_before_handle_nmi(&svm->vcpu);

	stgi();

	/* Any pending NMI will happen here */

	if (unlikely(svm->vmcb->control.exit_code == SVM_EXIT_NMI))
		kvm_after_handle_nmi(&svm->vcpu);

5005 5006
	sync_cr8_to_lapic(vcpu);

5007
	svm->next_rip = 0;
5008

5009 5010
	svm->vmcb->control.tlb_ctl = TLB_CONTROL_DO_NOTHING;

G
Gleb Natapov 已提交
5011 5012
	/* if exit due to PF check for async PF */
	if (svm->vmcb->control.exit_code == SVM_EXIT_EXCP_BASE + PF_VECTOR)
5013
		svm->vcpu.arch.apf.host_apf_reason = kvm_read_and_reset_pf_reason();
G
Gleb Natapov 已提交
5014

A
Avi Kivity 已提交
5015 5016 5017 5018
	if (npt_enabled) {
		vcpu->arch.regs_avail &= ~(1 << VCPU_EXREG_PDPTR);
		vcpu->arch.regs_dirty &= ~(1 << VCPU_EXREG_PDPTR);
	}
5019 5020 5021 5022 5023 5024 5025 5026

	/*
	 * We need to handle MC intercepts here before the vcpu has a chance to
	 * change the physical cpu
	 */
	if (unlikely(svm->vmcb->control.exit_code ==
		     SVM_EXIT_EXCP_BASE + MC_VECTOR))
		svm_handle_mce(svm);
5027 5028

	mark_all_clean(svm->vmcb);
A
Avi Kivity 已提交
5029
}
5030
STACK_FRAME_NON_STANDARD(svm_vcpu_run);
A
Avi Kivity 已提交
5031 5032 5033

static void svm_set_cr3(struct kvm_vcpu *vcpu, unsigned long root)
{
5034 5035
	struct vcpu_svm *svm = to_svm(vcpu);

5036
	svm->vmcb->save.cr3 = __sme_set(root);
5037
	mark_dirty(svm->vmcb, VMCB_CR);
5038
	svm_flush_tlb(vcpu);
A
Avi Kivity 已提交
5039 5040
}

5041 5042 5043 5044
static void set_tdp_cr3(struct kvm_vcpu *vcpu, unsigned long root)
{
	struct vcpu_svm *svm = to_svm(vcpu);

5045
	svm->vmcb->control.nested_cr3 = __sme_set(root);
5046
	mark_dirty(svm->vmcb, VMCB_NPT);
5047 5048

	/* Also sync guest cr3 here in case we live migrate */
5049
	svm->vmcb->save.cr3 = kvm_read_cr3(vcpu);
5050
	mark_dirty(svm->vmcb, VMCB_CR);
5051

5052
	svm_flush_tlb(vcpu);
5053 5054
}

A
Avi Kivity 已提交
5055 5056
static int is_disabled(void)
{
5057 5058 5059 5060 5061 5062
	u64 vm_cr;

	rdmsrl(MSR_VM_CR, vm_cr);
	if (vm_cr & (1 << SVM_VM_CR_SVM_DISABLE))
		return 1;

A
Avi Kivity 已提交
5063 5064 5065
	return 0;
}

I
Ingo Molnar 已提交
5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076
static void
svm_patch_hypercall(struct kvm_vcpu *vcpu, unsigned char *hypercall)
{
	/*
	 * Patch in the VMMCALL instruction:
	 */
	hypercall[0] = 0x0f;
	hypercall[1] = 0x01;
	hypercall[2] = 0xd9;
}

Y
Yang, Sheng 已提交
5077 5078 5079 5080 5081
static void svm_check_processor_compat(void *rtn)
{
	*(int *)rtn = 0;
}

5082 5083 5084 5085 5086
static bool svm_cpu_has_accelerated_tpr(void)
{
	return false;
}

5087 5088 5089 5090 5091
static bool svm_has_high_real_mode_segbase(void)
{
	return true;
}

5092 5093 5094 5095 5096
static u64 svm_get_mt_mask(struct kvm_vcpu *vcpu, gfn_t gfn, bool is_mmio)
{
	return 0;
}

5097 5098
static void svm_cpuid_update(struct kvm_vcpu *vcpu)
{
5099 5100 5101
	struct vcpu_svm *svm = to_svm(vcpu);

	/* Update nrips enabled cache */
5102
	svm->nrips_enabled = !!guest_cpuid_has(&svm->vcpu, X86_FEATURE_NRIPS);
5103 5104 5105 5106

	if (!kvm_vcpu_apicv_active(vcpu))
		return;

5107
	guest_cpuid_clear(vcpu, X86_FEATURE_X2APIC);
5108 5109
}

5110 5111
static void svm_set_supported_cpuid(u32 func, struct kvm_cpuid_entry2 *entry)
{
5112
	switch (func) {
5113 5114 5115 5116
	case 0x1:
		if (avic)
			entry->ecx &= ~bit(X86_FEATURE_X2APIC);
		break;
5117 5118 5119 5120
	case 0x80000001:
		if (nested)
			entry->ecx |= (1 << 2); /* Set SVM bit */
		break;
5121 5122 5123 5124 5125
	case 0x8000000A:
		entry->eax = 1; /* SVM revision 1 */
		entry->ebx = 8; /* Lets support 8 ASIDs in case we add proper
				   ASID emulation to nested SVM */
		entry->ecx = 0; /* Reserved */
5126 5127 5128 5129
		entry->edx = 0; /* Per default do not support any
				   additional features */

		/* Support next_rip if host supports it */
5130
		if (boot_cpu_has(X86_FEATURE_NRIPS))
5131
			entry->edx |= SVM_FEATURE_NRIP;
5132

5133 5134 5135 5136
		/* Support NPT for the guest if enabled */
		if (npt_enabled)
			entry->edx |= SVM_FEATURE_NPT;

5137 5138
		break;
	}
5139 5140
}

5141
static int svm_get_lpage_level(void)
5142
{
5143
	return PT_PDPE_LEVEL;
5144 5145
}

5146 5147
static bool svm_rdtscp_supported(void)
{
P
Paolo Bonzini 已提交
5148
	return boot_cpu_has(X86_FEATURE_RDTSCP);
5149 5150
}

5151 5152 5153 5154 5155
static bool svm_invpcid_supported(void)
{
	return false;
}

5156 5157 5158 5159 5160
static bool svm_mpx_supported(void)
{
	return false;
}

5161 5162 5163 5164 5165
static bool svm_xsaves_supported(void)
{
	return false;
}

5166 5167 5168 5169 5170
static bool svm_has_wbinvd_exit(void)
{
	return true;
}

5171
#define PRE_EX(exit)  { .exit_code = (exit), \
5172
			.stage = X86_ICPT_PRE_EXCEPT, }
5173
#define POST_EX(exit) { .exit_code = (exit), \
5174
			.stage = X86_ICPT_POST_EXCEPT, }
5175
#define POST_MEM(exit) { .exit_code = (exit), \
5176
			.stage = X86_ICPT_POST_MEMACCESS, }
5177

5178
static const struct __x86_intercept {
5179 5180 5181 5182 5183 5184 5185 5186
	u32 exit_code;
	enum x86_intercept_stage stage;
} x86_intercept_map[] = {
	[x86_intercept_cr_read]		= POST_EX(SVM_EXIT_READ_CR0),
	[x86_intercept_cr_write]	= POST_EX(SVM_EXIT_WRITE_CR0),
	[x86_intercept_clts]		= POST_EX(SVM_EXIT_WRITE_CR0),
	[x86_intercept_lmsw]		= POST_EX(SVM_EXIT_WRITE_CR0),
	[x86_intercept_smsw]		= POST_EX(SVM_EXIT_READ_CR0),
5187 5188
	[x86_intercept_dr_read]		= POST_EX(SVM_EXIT_READ_DR0),
	[x86_intercept_dr_write]	= POST_EX(SVM_EXIT_WRITE_DR0),
5189 5190 5191 5192 5193 5194 5195 5196
	[x86_intercept_sldt]		= POST_EX(SVM_EXIT_LDTR_READ),
	[x86_intercept_str]		= POST_EX(SVM_EXIT_TR_READ),
	[x86_intercept_lldt]		= POST_EX(SVM_EXIT_LDTR_WRITE),
	[x86_intercept_ltr]		= POST_EX(SVM_EXIT_TR_WRITE),
	[x86_intercept_sgdt]		= POST_EX(SVM_EXIT_GDTR_READ),
	[x86_intercept_sidt]		= POST_EX(SVM_EXIT_IDTR_READ),
	[x86_intercept_lgdt]		= POST_EX(SVM_EXIT_GDTR_WRITE),
	[x86_intercept_lidt]		= POST_EX(SVM_EXIT_IDTR_WRITE),
5197 5198 5199 5200 5201 5202 5203 5204
	[x86_intercept_vmrun]		= POST_EX(SVM_EXIT_VMRUN),
	[x86_intercept_vmmcall]		= POST_EX(SVM_EXIT_VMMCALL),
	[x86_intercept_vmload]		= POST_EX(SVM_EXIT_VMLOAD),
	[x86_intercept_vmsave]		= POST_EX(SVM_EXIT_VMSAVE),
	[x86_intercept_stgi]		= POST_EX(SVM_EXIT_STGI),
	[x86_intercept_clgi]		= POST_EX(SVM_EXIT_CLGI),
	[x86_intercept_skinit]		= POST_EX(SVM_EXIT_SKINIT),
	[x86_intercept_invlpga]		= POST_EX(SVM_EXIT_INVLPGA),
5205 5206 5207
	[x86_intercept_rdtscp]		= POST_EX(SVM_EXIT_RDTSCP),
	[x86_intercept_monitor]		= POST_MEM(SVM_EXIT_MONITOR),
	[x86_intercept_mwait]		= POST_EX(SVM_EXIT_MWAIT),
5208 5209 5210 5211 5212 5213 5214 5215 5216
	[x86_intercept_invlpg]		= POST_EX(SVM_EXIT_INVLPG),
	[x86_intercept_invd]		= POST_EX(SVM_EXIT_INVD),
	[x86_intercept_wbinvd]		= POST_EX(SVM_EXIT_WBINVD),
	[x86_intercept_wrmsr]		= POST_EX(SVM_EXIT_MSR),
	[x86_intercept_rdtsc]		= POST_EX(SVM_EXIT_RDTSC),
	[x86_intercept_rdmsr]		= POST_EX(SVM_EXIT_MSR),
	[x86_intercept_rdpmc]		= POST_EX(SVM_EXIT_RDPMC),
	[x86_intercept_cpuid]		= PRE_EX(SVM_EXIT_CPUID),
	[x86_intercept_rsm]		= PRE_EX(SVM_EXIT_RSM),
5217 5218 5219 5220 5221 5222 5223
	[x86_intercept_pause]		= PRE_EX(SVM_EXIT_PAUSE),
	[x86_intercept_pushf]		= PRE_EX(SVM_EXIT_PUSHF),
	[x86_intercept_popf]		= PRE_EX(SVM_EXIT_POPF),
	[x86_intercept_intn]		= PRE_EX(SVM_EXIT_SWINT),
	[x86_intercept_iret]		= PRE_EX(SVM_EXIT_IRET),
	[x86_intercept_icebp]		= PRE_EX(SVM_EXIT_ICEBP),
	[x86_intercept_hlt]		= POST_EX(SVM_EXIT_HLT),
5224 5225 5226 5227
	[x86_intercept_in]		= POST_EX(SVM_EXIT_IOIO),
	[x86_intercept_ins]		= POST_EX(SVM_EXIT_IOIO),
	[x86_intercept_out]		= POST_EX(SVM_EXIT_IOIO),
	[x86_intercept_outs]		= POST_EX(SVM_EXIT_IOIO),
5228 5229
};

5230
#undef PRE_EX
5231
#undef POST_EX
5232
#undef POST_MEM
5233

5234 5235 5236 5237
static int svm_check_intercept(struct kvm_vcpu *vcpu,
			       struct x86_instruction_info *info,
			       enum x86_intercept_stage stage)
{
5238 5239 5240 5241 5242 5243 5244 5245 5246 5247
	struct vcpu_svm *svm = to_svm(vcpu);
	int vmexit, ret = X86EMUL_CONTINUE;
	struct __x86_intercept icpt_info;
	struct vmcb *vmcb = svm->vmcb;

	if (info->intercept >= ARRAY_SIZE(x86_intercept_map))
		goto out;

	icpt_info = x86_intercept_map[info->intercept];

5248
	if (stage != icpt_info.stage)
5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262
		goto out;

	switch (icpt_info.exit_code) {
	case SVM_EXIT_READ_CR0:
		if (info->intercept == x86_intercept_cr_read)
			icpt_info.exit_code += info->modrm_reg;
		break;
	case SVM_EXIT_WRITE_CR0: {
		unsigned long cr0, val;
		u64 intercept;

		if (info->intercept == x86_intercept_cr_write)
			icpt_info.exit_code += info->modrm_reg;

5263 5264
		if (icpt_info.exit_code != SVM_EXIT_WRITE_CR0 ||
		    info->intercept == x86_intercept_clts)
5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287
			break;

		intercept = svm->nested.intercept;

		if (!(intercept & (1ULL << INTERCEPT_SELECTIVE_CR0)))
			break;

		cr0 = vcpu->arch.cr0 & ~SVM_CR0_SELECTIVE_MASK;
		val = info->src_val  & ~SVM_CR0_SELECTIVE_MASK;

		if (info->intercept == x86_intercept_lmsw) {
			cr0 &= 0xfUL;
			val &= 0xfUL;
			/* lmsw can't clear PE - catch this here */
			if (cr0 & X86_CR0_PE)
				val |= X86_CR0_PE;
		}

		if (cr0 ^ val)
			icpt_info.exit_code = SVM_EXIT_CR0_SEL_WRITE;

		break;
	}
5288 5289 5290 5291
	case SVM_EXIT_READ_DR0:
	case SVM_EXIT_WRITE_DR0:
		icpt_info.exit_code += info->modrm_reg;
		break;
5292 5293 5294 5295 5296 5297
	case SVM_EXIT_MSR:
		if (info->intercept == x86_intercept_wrmsr)
			vmcb->control.exit_info_1 = 1;
		else
			vmcb->control.exit_info_1 = 0;
		break;
5298 5299 5300 5301 5302 5303 5304
	case SVM_EXIT_PAUSE:
		/*
		 * We get this for NOP only, but pause
		 * is rep not, check this here
		 */
		if (info->rep_prefix != REPE_PREFIX)
			goto out;
5305 5306 5307 5308 5309 5310
	case SVM_EXIT_IOIO: {
		u64 exit_info;
		u32 bytes;

		if (info->intercept == x86_intercept_in ||
		    info->intercept == x86_intercept_ins) {
5311 5312
			exit_info = ((info->src_val & 0xffff) << 16) |
				SVM_IOIO_TYPE_MASK;
5313
			bytes = info->dst_bytes;
5314
		} else {
5315
			exit_info = (info->dst_val & 0xffff) << 16;
5316
			bytes = info->src_bytes;
5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336
		}

		if (info->intercept == x86_intercept_outs ||
		    info->intercept == x86_intercept_ins)
			exit_info |= SVM_IOIO_STR_MASK;

		if (info->rep_prefix)
			exit_info |= SVM_IOIO_REP_MASK;

		bytes = min(bytes, 4u);

		exit_info |= bytes << SVM_IOIO_SIZE_SHIFT;

		exit_info |= (u32)info->ad_bytes << (SVM_IOIO_ASIZE_SHIFT - 1);

		vmcb->control.exit_info_1 = exit_info;
		vmcb->control.exit_info_2 = info->next_rip;

		break;
	}
5337 5338 5339 5340
	default:
		break;
	}

5341 5342 5343
	/* TODO: Advertise NRIPS to guest hypervisor unconditionally */
	if (static_cpu_has(X86_FEATURE_NRIPS))
		vmcb->control.next_rip  = info->next_rip;
5344 5345 5346 5347 5348 5349 5350 5351
	vmcb->control.exit_code = icpt_info.exit_code;
	vmexit = nested_svm_exit_handled(svm);

	ret = (vmexit == NESTED_EXIT_DONE) ? X86EMUL_INTERCEPTED
					   : X86EMUL_CONTINUE;

out:
	return ret;
5352 5353
}

5354 5355 5356
static void svm_handle_external_intr(struct kvm_vcpu *vcpu)
{
	local_irq_enable();
P
Paolo Bonzini 已提交
5357 5358 5359 5360 5361 5362
	/*
	 * We must have an instruction with interrupts enabled, so
	 * the timer interrupt isn't delayed by the interrupt shadow.
	 */
	asm("nop");
	local_irq_disable();
5363 5364
}

5365 5366 5367 5368
static void svm_sched_in(struct kvm_vcpu *vcpu, int cpu)
{
}

5369 5370 5371 5372 5373 5374 5375 5376 5377
static inline void avic_post_state_restore(struct kvm_vcpu *vcpu)
{
	if (avic_handle_apic_id_update(vcpu) != 0)
		return;
	if (avic_handle_dfr_update(vcpu) != 0)
		return;
	avic_handle_ldr_update(vcpu);
}

5378 5379 5380 5381 5382 5383
static void svm_setup_mce(struct kvm_vcpu *vcpu)
{
	/* [63:9] are reserved. */
	vcpu->arch.mcg_cap &= 0x1ff;
}

5384
static struct kvm_x86_ops svm_x86_ops __ro_after_init = {
A
Avi Kivity 已提交
5385 5386 5387 5388
	.cpu_has_kvm_support = has_svm,
	.disabled_by_bios = is_disabled,
	.hardware_setup = svm_hardware_setup,
	.hardware_unsetup = svm_hardware_unsetup,
Y
Yang, Sheng 已提交
5389
	.check_processor_compatibility = svm_check_processor_compat,
A
Avi Kivity 已提交
5390 5391
	.hardware_enable = svm_hardware_enable,
	.hardware_disable = svm_hardware_disable,
5392
	.cpu_has_accelerated_tpr = svm_cpu_has_accelerated_tpr,
5393
	.cpu_has_high_real_mode_segbase = svm_has_high_real_mode_segbase,
A
Avi Kivity 已提交
5394 5395 5396

	.vcpu_create = svm_create_vcpu,
	.vcpu_free = svm_free_vcpu,
5397
	.vcpu_reset = svm_vcpu_reset,
A
Avi Kivity 已提交
5398

5399 5400 5401
	.vm_init = avic_vm_init,
	.vm_destroy = avic_vm_destroy,

5402
	.prepare_guest_switch = svm_prepare_guest_switch,
A
Avi Kivity 已提交
5403 5404
	.vcpu_load = svm_vcpu_load,
	.vcpu_put = svm_vcpu_put,
5405 5406
	.vcpu_blocking = svm_vcpu_blocking,
	.vcpu_unblocking = svm_vcpu_unblocking,
A
Avi Kivity 已提交
5407

5408
	.update_bp_intercept = update_bp_intercept,
A
Avi Kivity 已提交
5409 5410 5411 5412 5413
	.get_msr = svm_get_msr,
	.set_msr = svm_set_msr,
	.get_segment_base = svm_get_segment_base,
	.get_segment = svm_get_segment,
	.set_segment = svm_set_segment,
5414
	.get_cpl = svm_get_cpl,
5415
	.get_cs_db_l_bits = kvm_get_cs_db_l_bits,
5416
	.decache_cr0_guest_bits = svm_decache_cr0_guest_bits,
5417
	.decache_cr3 = svm_decache_cr3,
5418
	.decache_cr4_guest_bits = svm_decache_cr4_guest_bits,
A
Avi Kivity 已提交
5419 5420 5421 5422 5423 5424 5425 5426
	.set_cr0 = svm_set_cr0,
	.set_cr3 = svm_set_cr3,
	.set_cr4 = svm_set_cr4,
	.set_efer = svm_set_efer,
	.get_idt = svm_get_idt,
	.set_idt = svm_set_idt,
	.get_gdt = svm_get_gdt,
	.set_gdt = svm_set_gdt,
J
Jan Kiszka 已提交
5427 5428
	.get_dr6 = svm_get_dr6,
	.set_dr6 = svm_set_dr6,
5429
	.set_dr7 = svm_set_dr7,
5430
	.sync_dirty_debug_regs = svm_sync_dirty_debug_regs,
A
Avi Kivity 已提交
5431
	.cache_reg = svm_cache_reg,
A
Avi Kivity 已提交
5432 5433
	.get_rflags = svm_get_rflags,
	.set_rflags = svm_set_rflags,
5434

A
Avi Kivity 已提交
5435 5436 5437
	.tlb_flush = svm_flush_tlb,

	.run = svm_vcpu_run,
5438
	.handle_exit = handle_exit,
A
Avi Kivity 已提交
5439
	.skip_emulated_instruction = skip_emulated_instruction,
5440 5441
	.set_interrupt_shadow = svm_set_interrupt_shadow,
	.get_interrupt_shadow = svm_get_interrupt_shadow,
I
Ingo Molnar 已提交
5442
	.patch_hypercall = svm_patch_hypercall,
E
Eddie Dong 已提交
5443
	.set_irq = svm_set_irq,
5444
	.set_nmi = svm_inject_nmi,
5445
	.queue_exception = svm_queue_exception,
A
Avi Kivity 已提交
5446
	.cancel_injection = svm_cancel_injection,
5447
	.interrupt_allowed = svm_interrupt_allowed,
5448
	.nmi_allowed = svm_nmi_allowed,
J
Jan Kiszka 已提交
5449 5450
	.get_nmi_mask = svm_get_nmi_mask,
	.set_nmi_mask = svm_set_nmi_mask,
5451 5452 5453
	.enable_nmi_window = enable_nmi_window,
	.enable_irq_window = enable_irq_window,
	.update_cr8_intercept = update_cr8_intercept,
5454
	.set_virtual_x2apic_mode = svm_set_virtual_x2apic_mode,
5455 5456
	.get_enable_apicv = svm_get_enable_apicv,
	.refresh_apicv_exec_ctrl = svm_refresh_apicv_exec_ctrl,
5457
	.load_eoi_exitmap = svm_load_eoi_exitmap,
5458 5459
	.hwapic_irr_update = svm_hwapic_irr_update,
	.hwapic_isr_update = svm_hwapic_isr_update,
5460
	.apicv_post_state_restore = avic_post_state_restore,
5461 5462

	.set_tss_addr = svm_set_tss_addr,
5463
	.get_tdp_level = get_npt_level,
5464
	.get_mt_mask = svm_get_mt_mask,
5465

5466 5467
	.get_exit_info = svm_get_exit_info,

5468
	.get_lpage_level = svm_get_lpage_level,
5469 5470

	.cpuid_update = svm_cpuid_update,
5471 5472

	.rdtscp_supported = svm_rdtscp_supported,
5473
	.invpcid_supported = svm_invpcid_supported,
5474
	.mpx_supported = svm_mpx_supported,
5475
	.xsaves_supported = svm_xsaves_supported,
5476 5477

	.set_supported_cpuid = svm_set_supported_cpuid,
5478 5479

	.has_wbinvd_exit = svm_has_wbinvd_exit,
5480 5481

	.write_tsc_offset = svm_write_tsc_offset,
5482 5483

	.set_tdp_cr3 = set_tdp_cr3,
5484 5485

	.check_intercept = svm_check_intercept,
5486
	.handle_external_intr = svm_handle_external_intr,
5487 5488

	.sched_in = svm_sched_in,
5489 5490

	.pmu_ops = &amd_pmu_ops,
5491
	.deliver_posted_interrupt = svm_deliver_avic_intr,
5492
	.update_pi_irte = svm_update_pi_irte,
5493
	.setup_mce = svm_setup_mce,
A
Avi Kivity 已提交
5494 5495 5496 5497
};

static int __init svm_init(void)
{
5498
	return kvm_init(&svm_x86_ops, sizeof(struct vcpu_svm),
5499
			__alignof__(struct vcpu_svm), THIS_MODULE);
A
Avi Kivity 已提交
5500 5501 5502 5503
}

static void __exit svm_exit(void)
{
5504
	kvm_exit();
A
Avi Kivity 已提交
5505 5506 5507 5508
}

module_init(svm_init)
module_exit(svm_exit)