svm.c 175.6 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>
B
Brijesh Singh 已提交
40
#include <linux/psp-sev.h>
B
Brijesh Singh 已提交
41
#include <linux/file.h>
42 43
#include <linux/pagemap.h>
#include <linux/swap.h>
A
Avi Kivity 已提交
44

45
#include <asm/apic.h>
46
#include <asm/perf_event.h>
47
#include <asm/tlbflush.h>
A
Avi Kivity 已提交
48
#include <asm/desc.h>
49
#include <asm/debugreg.h>
G
Gleb Natapov 已提交
50
#include <asm/kvm_para.h>
51
#include <asm/irq_remapping.h>
52
#include <asm/nospec-branch.h>
A
Avi Kivity 已提交
53

54
#include <asm/virtext.h>
55
#include "trace.h"
56

57 58
#define __ex(x) __kvm_handle_fault_on_reboot(x)

A
Avi Kivity 已提交
59 60 61
MODULE_AUTHOR("Qumranet");
MODULE_LICENSE("GPL");

62 63 64 65 66 67
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
68 69 70 71 72 73
#define IOPM_ALLOC_ORDER 2
#define MSRPM_ALLOC_ORDER 1

#define SEG_TYPE_LDT 2
#define SEG_TYPE_BUSY_TSS16 3

74 75 76 77
#define SVM_FEATURE_NPT            (1 <<  0)
#define SVM_FEATURE_LBRV           (1 <<  1)
#define SVM_FEATURE_SVML           (1 <<  2)
#define SVM_FEATURE_NRIP           (1 <<  3)
78 79 80 81
#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)
82
#define SVM_FEATURE_PAUSE_FILTER   (1 << 10)
83

84 85
#define SVM_AVIC_DOORBELL	0xc001011b

86 87 88 89
#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      */

90 91
#define DEBUGCTL_RESERVED_BITS (~(0x3fULL))

92
#define TSC_RATIO_RSVD          0xffffff0000000000ULL
93 94
#define TSC_RATIO_MIN		0x0000000000000001ULL
#define TSC_RATIO_MAX		0x000000ffffffffffULL
95

D
Dan Carpenter 已提交
96
#define AVIC_HPA_MASK	~((0xFFFULL << 52) | 0xFFF)
97 98 99 100 101 102 103

/*
 * 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

104 105 106 107
#define AVIC_UNACCEL_ACCESS_WRITE_MASK		1
#define AVIC_UNACCEL_ACCESS_OFFSET_MASK		0xFF0
#define AVIC_UNACCEL_ACCESS_VECTOR_MASK		0xFFFFFFFF

108 109 110 111 112 113 114 115 116 117 118 119 120
/* 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)

121 122
static bool erratum_383_found __read_mostly;

A
Avi Kivity 已提交
123 124 125 126 127 128
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
129
	MSR_TSC_AUX,
A
Avi Kivity 已提交
130 131 132 133 134 135
};

#define NR_HOST_SAVE_USER_MSRS ARRAY_SIZE(host_save_user_msrs)

struct kvm_vcpu;

136 137 138
struct nested_state {
	struct vmcb *hsave;
	u64 hsave_msr;
139
	u64 vm_cr_msr;
140 141 142 143 144 145 146
	u64 vmcb;

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

	/* gpa pointers to the real vectors */
	u64 vmcb_msrpm;
147
	u64 vmcb_iopm;
J
Joerg Roedel 已提交
148

149 150 151
	/* A VMEXIT is required but not yet emulated */
	bool exit_required;

J
Joerg Roedel 已提交
152
	/* cache for intercepts of the guest */
153
	u32 intercept_cr;
154
	u32 intercept_dr;
J
Joerg Roedel 已提交
155 156 157
	u32 intercept_exceptions;
	u64 intercept;

158 159
	/* Nested Paging related state */
	u64 nested_cr3;
160 161
};

162 163 164
#define MSRPM_OFFSETS	16
static u32 msrpm_offsets[MSRPM_OFFSETS] __read_mostly;

165 166 167 168 169 170
/*
 * 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
171 172 173 174 175 176 177 178
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
179
	uint64_t tsc_aux;
A
Avi Kivity 已提交
180

181 182
	u64 msr_decfg;

A
Avi Kivity 已提交
183 184 185
	u64 next_rip;

	u64 host_user_msrs[NR_HOST_SAVE_USER_MSRS];
186
	struct {
187 188 189
		u16 fs;
		u16 gs;
		u16 ldt;
190 191
		u64 gs_base;
	} host;
A
Avi Kivity 已提交
192

193 194
	u64 spec_ctrl;

A
Avi Kivity 已提交
195 196
	u32 *msrpm;

197 198
	ulong nmi_iret_rip;

199
	struct nested_state nested;
J
Jan Kiszka 已提交
200 201

	bool nmi_singlestep;
202
	u64 nmi_singlestep_guest_rflags;
203 204 205

	unsigned int3_injected;
	unsigned long int3_rip;
206

207 208
	/* cached guest cpuid flags for faster access */
	bool nrips_enabled	: 1;
209

210
	u32 ldr_reg;
211 212
	struct page *avic_backing_page;
	u64 *avic_physical_id_cache;
213
	bool avic_is_running;
214 215 216 217 218 219 220 221 222

	/*
	 * 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;
223 224 225

	/* which host CPU was used for running this vcpu */
	unsigned int last_cpu;
226 227 228 229 230 231 232 233
};

/*
 * 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
234 235
};

236 237 238 239 240 241 242 243
#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)

244 245 246
static DEFINE_PER_CPU(u64, current_tsc_ratio);
#define TSC_RATIO_DEFAULT	0x0100000000ULL

247 248
#define MSR_INVALID			0xffffffffU

249
static const struct svm_direct_access_msrs {
250 251 252
	u32 index;   /* Index of the MSR */
	bool always; /* True if intercept is always on */
} direct_access_msrs[] = {
B
Brian Gerst 已提交
253
	{ .index = MSR_STAR,				.always = true  },
254 255 256 257 258 259 260 261 262
	{ .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
263
	{ .index = MSR_IA32_SPEC_CTRL,			.always = false },
A
Ashok Raj 已提交
264
	{ .index = MSR_IA32_PRED_CMD,			.always = false },
265 266 267 268 269
	{ .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
270 271
};

272 273 274 275
/* 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
276
static bool npt_enabled;
277
#endif
278

279 280
/* allow nested paging (virtualized MMU) for all guests */
static int npt = true;
281
module_param(npt, int, S_IRUGO);
282

283 284
/* allow nested virtualization in KVM/SVM */
static int nested = true;
285 286
module_param(nested, int, S_IRUGO);

287 288
/* enable / disable AVIC */
static int avic;
289
#ifdef CONFIG_X86_LOCAL_APIC
290
module_param(avic, int, S_IRUGO);
291
#endif
292

293 294 295 296
/* enable/disable Virtual VMLOAD VMSAVE */
static int vls = true;
module_param(vls, int, 0444);

297 298 299
/* enable/disable Virtual GIF */
static int vgif = true;
module_param(vgif, int, 0444);
300

B
Brijesh Singh 已提交
301 302 303 304
/* enable/disable SEV support */
static int sev = IS_ENABLED(CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT);
module_param(sev, int, 0444);

B
Brijesh Singh 已提交
305 306
static u8 rsm_ins_bytes[] = "\x0f\xaa";

307
static void svm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0);
308
static void svm_flush_tlb(struct kvm_vcpu *vcpu, bool invalidate_gpa);
309
static void svm_complete_interrupts(struct vcpu_svm *svm);
310

311
static int nested_svm_exit_handled(struct vcpu_svm *svm);
312
static int nested_svm_intercept(struct vcpu_svm *svm);
313 314 315 316
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);

317
enum {
318 319
	VMCB_INTERCEPTS, /* Intercept vectors, TSC offset,
			    pause filter count */
320
	VMCB_PERM_MAP,   /* IOPM Base and MSRPM Base */
321
	VMCB_ASID,	 /* ASID */
322
	VMCB_INTR,	 /* int_ctl, int_vector */
323
	VMCB_NPT,        /* npt_en, nCR3, gPAT */
324
	VMCB_CR,	 /* CR0, CR3, CR4, EFER */
325
	VMCB_DR,         /* DR6, DR7 */
326
	VMCB_DT,         /* GDT, IDT */
327
	VMCB_SEG,        /* CS, DS, SS, ES, CPL */
328
	VMCB_CR2,        /* CR2 only */
329
	VMCB_LBR,        /* DBGCTL, BR_FROM, BR_TO, LAST_EX_FROM, LAST_EX_TO */
330 331 332 333
	VMCB_AVIC,       /* AVIC APIC_BAR, AVIC APIC_BACKING_PAGE,
			  * AVIC PHYSICAL_TABLE pointer,
			  * AVIC LOGICAL_TABLE pointer
			  */
334 335 336
	VMCB_DIRTY_MAX,
};

337 338
/* TPR and CR2 are always written before VMRUN */
#define VMCB_ALWAYS_DIRTY_MASK	((1U << VMCB_INTR) | (1U << VMCB_CR2))
339

340 341
#define VMCB_AVIC_APIC_BAR_MASK		0xFFFFFFFFFF000ULL

342
static unsigned int max_sev_asid;
B
Brijesh Singh 已提交
343 344
static unsigned int min_sev_asid;
static unsigned long *sev_asid_bitmap;
345
#define __sme_page_pa(x) __sme_set(page_to_pfn(x) << PAGE_SHIFT)
B
Brijesh Singh 已提交
346

347 348 349 350 351 352 353 354
struct enc_region {
	struct list_head list;
	unsigned long npages;
	struct page **pages;
	unsigned long uaddr;
	unsigned long size;
};

B
Brijesh Singh 已提交
355 356 357 358 359 360 361 362 363 364 365
static inline bool svm_sev_enabled(void)
{
	return max_sev_asid;
}

static inline bool sev_guest(struct kvm *kvm)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;

	return sev->active;
}
366

367 368 369 370 371 372 373
static inline int sev_get_asid(struct kvm *kvm)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;

	return sev->asid;
}

374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389
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);
}

390 391
static inline struct vcpu_svm *to_svm(struct kvm_vcpu *vcpu)
{
R
Rusty Russell 已提交
392
	return container_of(vcpu, struct vcpu_svm, vcpu);
393 394
}

395 396 397 398 399 400
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);
}

401 402 403 404 405 406 407 408 409 410 411
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);
}

412 413 414 415 416
static void recalc_intercepts(struct vcpu_svm *svm)
{
	struct vmcb_control_area *c, *h;
	struct nested_state *g;

417 418
	mark_dirty(svm->vmcb, VMCB_INTERCEPTS);

419 420 421 422 423 424 425
	if (!is_guest_mode(&svm->vcpu))
		return;

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

426
	c->intercept_cr = h->intercept_cr | g->intercept_cr;
427
	c->intercept_dr = h->intercept_dr | g->intercept_dr;
428
	c->intercept_exceptions = h->intercept_exceptions | g->intercept_exceptions;
429 430 431
	c->intercept = h->intercept | g->intercept;
}

432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
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);
}

465
static inline void set_dr_intercepts(struct vcpu_svm *svm)
466 467 468
{
	struct vmcb *vmcb = get_host_vmcb(svm);

469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484
	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);
485 486 487 488

	recalc_intercepts(svm);
}

489
static inline void clr_dr_intercepts(struct vcpu_svm *svm)
490 491 492
{
	struct vmcb *vmcb = get_host_vmcb(svm);

493
	vmcb->control.intercept_dr = 0;
494 495 496 497

	recalc_intercepts(svm);
}

498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
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);
}

516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
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);
}

534 535 536 537 538
static inline bool vgif_enabled(struct vcpu_svm *svm)
{
	return !!(svm->vmcb->control.int_ctl & V_GIF_ENABLE_MASK);
}

539 540
static inline void enable_gif(struct vcpu_svm *svm)
{
541 542 543 544
	if (vgif_enabled(svm))
		svm->vmcb->control.int_ctl |= V_GIF_MASK;
	else
		svm->vcpu.arch.hflags |= HF_GIF_MASK;
545 546 547 548
}

static inline void disable_gif(struct vcpu_svm *svm)
{
549 550 551 552
	if (vgif_enabled(svm))
		svm->vmcb->control.int_ctl &= ~V_GIF_MASK;
	else
		svm->vcpu.arch.hflags &= ~HF_GIF_MASK;
553 554 555 556
}

static inline bool gif_set(struct vcpu_svm *svm)
{
557 558 559 560
	if (vgif_enabled(svm))
		return !!(svm->vmcb->control.int_ctl & V_GIF_MASK);
	else
		return !!(svm->vcpu.arch.hflags & HF_GIF_MASK);
561 562
}

563
static unsigned long iopm_base;
A
Avi Kivity 已提交
564 565 566 567

struct kvm_ldttss_desc {
	u16 limit0;
	u16 base0;
J
Joerg Roedel 已提交
568 569
	unsigned base1:8, type:5, dpl:2, p:1;
	unsigned limit1:4, zero0:3, g:1, base2:8;
A
Avi Kivity 已提交
570 571 572 573 574 575 576
	u32 base3;
	u32 zero1;
} __attribute__((packed));

struct svm_cpu_data {
	int cpu;

A
Avi Kivity 已提交
577 578 579
	u64 asid_generation;
	u32 max_asid;
	u32 next_asid;
580
	u32 min_asid;
A
Avi Kivity 已提交
581 582 583
	struct kvm_ldttss_desc *tss_desc;

	struct page *save_area;
A
Ashok Raj 已提交
584
	struct vmcb *current_vmcb;
585 586 587

	/* index = sev_asid, value = vmcb pointer */
	struct vmcb **sev_vmcbs;
A
Avi Kivity 已提交
588 589 590 591 592 593 594 595 596
};

static DEFINE_PER_CPU(struct svm_cpu_data *, svm_data);

struct svm_init_data {
	int cpu;
	int r;
};

597
static const u32 msrpm_ranges[] = {0, 0xc0000000, 0xc0010000};
A
Avi Kivity 已提交
598

599
#define NUM_MSR_MAPS ARRAY_SIZE(msrpm_ranges)
A
Avi Kivity 已提交
600 601 602
#define MSRS_RANGE_SIZE 2048
#define MSRS_IN_RANGE (MSRS_RANGE_SIZE * 8 / 2)

603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623
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
624 625 626 627
#define MAX_INST_SIZE 15

static inline void clgi(void)
{
628
	asm volatile (__ex(SVM_CLGI));
A
Avi Kivity 已提交
629 630 631 632
}

static inline void stgi(void)
{
633
	asm volatile (__ex(SVM_STGI));
A
Avi Kivity 已提交
634 635 636 637
}

static inline void invlpga(unsigned long addr, u32 asid)
{
J
Joerg Roedel 已提交
638
	asm volatile (__ex(SVM_INVLPGA) : : "a"(addr), "c"(asid));
A
Avi Kivity 已提交
639 640
}

641
static int get_npt_level(struct kvm_vcpu *vcpu)
642 643
{
#ifdef CONFIG_X86_64
644
	return PT64_ROOT_4LEVEL;
645 646 647 648 649
#else
	return PT32E_ROOT_LEVEL;
#endif
}

A
Avi Kivity 已提交
650 651
static void svm_set_efer(struct kvm_vcpu *vcpu, u64 efer)
{
652
	vcpu->arch.efer = efer;
653
	if (!npt_enabled && !(efer & EFER_LMA))
654
		efer &= ~EFER_LME;
A
Avi Kivity 已提交
655

656
	to_svm(vcpu)->vmcb->save.efer = efer | EFER_SVME;
657
	mark_dirty(to_svm(vcpu)->vmcb, VMCB_CR);
A
Avi Kivity 已提交
658 659 660 661 662 663 664 665
}

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

666
static u32 svm_get_interrupt_shadow(struct kvm_vcpu *vcpu)
667 668 669 670 671
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u32 ret = 0;

	if (svm->vmcb->control.int_state & SVM_INTERRUPT_SHADOW_MASK)
672 673
		ret = KVM_X86_SHADOW_INT_STI | KVM_X86_SHADOW_INT_MOV_SS;
	return ret;
674 675 676 677 678 679 680 681 682 683 684 685 686
}

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
687 688
static void skip_emulated_instruction(struct kvm_vcpu *vcpu)
{
689 690
	struct vcpu_svm *svm = to_svm(vcpu);

691
	if (svm->vmcb->control.next_rip != 0) {
692
		WARN_ON_ONCE(!static_cpu_has(X86_FEATURE_NRIPS));
693
		svm->next_rip = svm->vmcb->control.next_rip;
694
	}
695

696
	if (!svm->next_rip) {
697
		if (emulate_instruction(vcpu, EMULTYPE_SKIP) !=
698 699
				EMULATE_DONE)
			printk(KERN_DEBUG "%s: NOP\n", __func__);
A
Avi Kivity 已提交
700 701
		return;
	}
702 703 704
	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
705

706
	kvm_rip_write(vcpu, svm->next_rip);
707
	svm_set_interrupt_shadow(vcpu, 0);
A
Avi Kivity 已提交
708 709
}

710
static void svm_queue_exception(struct kvm_vcpu *vcpu)
J
Jan Kiszka 已提交
711 712
{
	struct vcpu_svm *svm = to_svm(vcpu);
713 714
	unsigned nr = vcpu->arch.exception.nr;
	bool has_error_code = vcpu->arch.exception.has_error_code;
715
	bool reinject = vcpu->arch.exception.injected;
716
	u32 error_code = vcpu->arch.exception.error_code;
J
Jan Kiszka 已提交
717

J
Joerg Roedel 已提交
718 719 720 721
	/*
	 * If we are within a nested VM we'd better #VMEXIT and let the guest
	 * handle the exception
	 */
722 723
	if (!reinject &&
	    nested_svm_check_exception(svm, nr, has_error_code, error_code))
J
Jan Kiszka 已提交
724 725
		return;

726
	if (nr == BP_VECTOR && !static_cpu_has(X86_FEATURE_NRIPS)) {
727 728 729 730 731 732 733 734 735 736 737 738 739 740 741
		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
742 743 744 745 746 747 748
	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;
}

749 750 751 752 753 754
static void svm_init_erratum_383(void)
{
	u32 low, high;
	int err;
	u64 val;

755
	if (!static_cpu_has_bug(X86_BUG_AMD_TLB_MMATCH))
756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772
		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;
}

773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793
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
794 795
static int has_svm(void)
{
796
	const char *msg;
A
Avi Kivity 已提交
797

798
	if (!cpu_has_svm(&msg)) {
J
Joe Perches 已提交
799
		printk(KERN_INFO "has_svm: %s\n", msg);
A
Avi Kivity 已提交
800 801 802 803 804 805
		return 0;
	}

	return 1;
}

806
static void svm_hardware_disable(void)
A
Avi Kivity 已提交
807
{
808 809 810 811
	/* Make sure we clean up behind us */
	if (static_cpu_has(X86_FEATURE_TSCRATEMSR))
		wrmsrl(MSR_AMD64_TSC_RATIO, TSC_RATIO_DEFAULT);

812
	cpu_svm_disable();
813 814

	amd_pmu_disable_virt();
A
Avi Kivity 已提交
815 816
}

817
static int svm_hardware_enable(void)
A
Avi Kivity 已提交
818 819
{

820
	struct svm_cpu_data *sd;
A
Avi Kivity 已提交
821 822 823 824
	uint64_t efer;
	struct desc_struct *gdt;
	int me = raw_smp_processor_id();

825 826 827 828
	rdmsrl(MSR_EFER, efer);
	if (efer & EFER_SVME)
		return -EBUSY;

A
Avi Kivity 已提交
829
	if (!has_svm()) {
830
		pr_err("%s: err EOPNOTSUPP on %d\n", __func__, me);
831
		return -EINVAL;
A
Avi Kivity 已提交
832
	}
833 834
	sd = per_cpu(svm_data, me);
	if (!sd) {
835
		pr_err("%s: svm_data is NULL on %d\n", __func__, me);
836
		return -EINVAL;
A
Avi Kivity 已提交
837 838
	}

839 840 841
	sd->asid_generation = 1;
	sd->max_asid = cpuid_ebx(SVM_CPUID_FUNC) - 1;
	sd->next_asid = sd->max_asid + 1;
842
	sd->min_asid = max_sev_asid + 1;
A
Avi Kivity 已提交
843

844
	gdt = get_current_gdt_rw();
845
	sd->tss_desc = (struct kvm_ldttss_desc *)(gdt + GDT_ENTRY_TSS);
A
Avi Kivity 已提交
846

847
	wrmsrl(MSR_EFER, efer | EFER_SVME);
A
Avi Kivity 已提交
848

849
	wrmsrl(MSR_VM_HSAVE_PA, page_to_pfn(sd->save_area) << PAGE_SHIFT);
850

851 852
	if (static_cpu_has(X86_FEATURE_TSCRATEMSR)) {
		wrmsrl(MSR_AMD64_TSC_RATIO, TSC_RATIO_DEFAULT);
853
		__this_cpu_write(current_tsc_ratio, TSC_RATIO_DEFAULT);
854 855
	}

856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885

	/*
	 * 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;

886 887
	svm_init_erratum_383();

888 889
	amd_pmu_enable_virt();

890
	return 0;
A
Avi Kivity 已提交
891 892
}

893 894
static void svm_cpu_uninit(int cpu)
{
895
	struct svm_cpu_data *sd = per_cpu(svm_data, raw_smp_processor_id());
896

897
	if (!sd)
898 899 900
		return;

	per_cpu(svm_data, raw_smp_processor_id()) = NULL;
901
	kfree(sd->sev_vmcbs);
902 903
	__free_page(sd->save_area);
	kfree(sd);
904 905
}

A
Avi Kivity 已提交
906 907
static int svm_cpu_init(int cpu)
{
908
	struct svm_cpu_data *sd;
A
Avi Kivity 已提交
909 910
	int r;

911 912
	sd = kzalloc(sizeof(struct svm_cpu_data), GFP_KERNEL);
	if (!sd)
A
Avi Kivity 已提交
913
		return -ENOMEM;
914
	sd->cpu = cpu;
A
Avi Kivity 已提交
915
	r = -ENOMEM;
916
	sd->save_area = alloc_page(GFP_KERNEL);
917
	if (!sd->save_area)
A
Avi Kivity 已提交
918 919
		goto err_1;

920 921 922 923 924 925 926
	if (svm_sev_enabled()) {
		r = -ENOMEM;
		sd->sev_vmcbs = kmalloc((max_sev_asid + 1) * sizeof(void *), GFP_KERNEL);
		if (!sd->sev_vmcbs)
			goto err_1;
	}

927
	per_cpu(svm_data, cpu) = sd;
A
Avi Kivity 已提交
928 929 930 931

	return 0;

err_1:
932
	kfree(sd);
A
Avi Kivity 已提交
933 934 935 936
	return r;

}

937 938 939 940 941 942 943 944 945 946 947
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;
}

948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966
static bool msr_write_intercepted(struct kvm_vcpu *vcpu, unsigned msr)
{
	u8 bit_write;
	unsigned long tmp;
	u32 offset;
	u32 *msrpm;

	msrpm = is_guest_mode(vcpu) ? to_svm(vcpu)->nested.msrpm:
				      to_svm(vcpu)->msrpm;

	offset    = svm_msrpm_offset(msr);
	bit_write = 2 * (msr & 0x0f) + 1;
	tmp       = msrpm[offset];

	BUG_ON(offset == MSR_INVALID);

	return !!test_bit(bit_write,  &tmp);
}

967 968
static void set_msr_interception(u32 *msrpm, unsigned msr,
				 int read, int write)
A
Avi Kivity 已提交
969
{
970 971 972
	u8 bit_read, bit_write;
	unsigned long tmp;
	u32 offset;
A
Avi Kivity 已提交
973

974 975 976 977 978 979
	/*
	 * If this warning triggers extend the direct_access_msrs list at the
	 * beginning of the file
	 */
	WARN_ON(!valid_msr_intercept(msr));

980 981 982 983 984 985 986 987 988 989 990
	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
991 992
}

993
static void svm_vcpu_init_msrpm(u32 *msrpm)
A
Avi Kivity 已提交
994 995 996
{
	int i;

997 998
	memset(msrpm, 0xff, PAGE_SIZE * (1 << MSRPM_ALLOC_ORDER));

999 1000 1001 1002 1003 1004
	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);
	}
1005 1006
}

1007 1008 1009 1010 1011 1012 1013 1014
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)
1015
			return;
1016 1017 1018 1019 1020 1021 1022 1023 1024

		/* 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
1025
	}
1026 1027 1028 1029 1030

	/*
	 * If this BUG triggers the msrpm_offsets table has an overflow. Just
	 * increase MSRPM_OFFSETS in this case.
	 */
1031
	BUG();
A
Avi Kivity 已提交
1032 1033
}

1034
static void init_msrpm_offsets(void)
1035
{
1036
	int i;
1037

1038 1039 1040 1041 1042 1043 1044 1045 1046 1047
	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);
	}
1048 1049
}

1050 1051 1052 1053
static void svm_enable_lbrv(struct vcpu_svm *svm)
{
	u32 *msrpm = svm->msrpm;

1054
	svm->vmcb->control.virt_ext |= LBR_CTL_ENABLE_MASK;
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
	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;

1065
	svm->vmcb->control.virt_ext &= ~LBR_CTL_ENABLE_MASK;
1066 1067 1068 1069 1070 1071
	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);
}

1072 1073 1074
static void disable_nmi_singlestep(struct vcpu_svm *svm)
{
	svm->nmi_singlestep = false;
1075

1076 1077 1078 1079 1080 1081 1082
	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;
	}
1083 1084
}

1085 1086 1087 1088 1089 1090
/* 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
1091
static DEFINE_HASHTABLE(svm_vm_data_hash, SVM_VM_DATA_HASH_BITS);
1092 1093
static u32 next_vm_id = 0;
static bool next_vm_id_wrapped = 0;
1094
static DEFINE_SPINLOCK(svm_vm_data_hash_lock);
1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126

/* 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);

	/* 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.
	 */
1127
	if (vcpu)
1128 1129 1130 1131 1132
		kvm_vcpu_wake_up(vcpu);

	return 0;
}

B
Brijesh Singh 已提交
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
static __init int sev_hardware_setup(void)
{
	struct sev_user_data_status *status;
	int rc;

	/* Maximum number of encrypted guests supported simultaneously */
	max_sev_asid = cpuid_ecx(0x8000001F);

	if (!max_sev_asid)
		return 1;

B
Brijesh Singh 已提交
1144 1145 1146 1147 1148 1149 1150 1151 1152
	/* Minimum ASID value that should be used for SEV guest */
	min_sev_asid = cpuid_edx(0x8000001F);

	/* Initialize SEV ASID bitmap */
	sev_asid_bitmap = kcalloc(BITS_TO_LONGS(max_sev_asid),
				sizeof(unsigned long), GFP_KERNEL);
	if (!sev_asid_bitmap)
		return 1;

B
Brijesh Singh 已提交
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174
	status = kmalloc(sizeof(*status), GFP_KERNEL);
	if (!status)
		return 1;

	/*
	 * Check SEV platform status.
	 *
	 * PLATFORM_STATUS can be called in any state, if we failed to query
	 * the PLATFORM status then either PSP firmware does not support SEV
	 * feature or SEV firmware is dead.
	 */
	rc = sev_platform_status(status, NULL);
	if (rc)
		goto err;

	pr_info("SEV supported\n");

err:
	kfree(status);
	return rc;
}

A
Avi Kivity 已提交
1175 1176 1177 1178
static __init int svm_hardware_setup(void)
{
	int cpu;
	struct page *iopm_pages;
1179
	void *iopm_va;
A
Avi Kivity 已提交
1180 1181 1182 1183 1184 1185
	int r;

	iopm_pages = alloc_pages(GFP_KERNEL, IOPM_ALLOC_ORDER);

	if (!iopm_pages)
		return -ENOMEM;
1186 1187 1188

	iopm_va = page_address(iopm_pages);
	memset(iopm_va, 0xff, PAGE_SIZE * (1 << IOPM_ALLOC_ORDER));
A
Avi Kivity 已提交
1189 1190
	iopm_base = page_to_pfn(iopm_pages) << PAGE_SHIFT;

1191 1192
	init_msrpm_offsets();

1193 1194 1195
	if (boot_cpu_has(X86_FEATURE_NX))
		kvm_enable_efer_bits(EFER_NX);

A
Alexander Graf 已提交
1196 1197 1198
	if (boot_cpu_has(X86_FEATURE_FXSR_OPT))
		kvm_enable_efer_bits(EFER_FFXSR);

1199 1200
	if (boot_cpu_has(X86_FEATURE_TSCRATEMSR)) {
		kvm_has_tsc_control = true;
1201 1202
		kvm_max_tsc_scaling_ratio = TSC_RATIO_MAX;
		kvm_tsc_scaling_ratio_frac_bits = 32;
1203 1204
	}

1205 1206
	if (nested) {
		printk(KERN_INFO "kvm: Nested Virtualization enabled\n");
1207
		kvm_enable_efer_bits(EFER_SVME | EFER_LMSLE);
1208 1209
	}

B
Brijesh Singh 已提交
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220
	if (sev) {
		if (boot_cpu_has(X86_FEATURE_SEV) &&
		    IS_ENABLED(CONFIG_KVM_AMD_SEV)) {
			r = sev_hardware_setup();
			if (r)
				sev = false;
		} else {
			sev = false;
		}
	}

Z
Zachary Amsden 已提交
1221
	for_each_possible_cpu(cpu) {
A
Avi Kivity 已提交
1222 1223
		r = svm_cpu_init(cpu);
		if (r)
1224
			goto err;
A
Avi Kivity 已提交
1225
	}
1226

1227
	if (!boot_cpu_has(X86_FEATURE_NPT))
1228 1229
		npt_enabled = false;

1230 1231 1232 1233 1234
	if (npt_enabled && !npt) {
		printk(KERN_INFO "kvm: Nested Paging disabled\n");
		npt_enabled = false;
	}

1235
	if (npt_enabled) {
1236
		printk(KERN_INFO "kvm: Nested Paging enabled\n");
1237
		kvm_enable_tdp();
1238 1239
	} else
		kvm_disable_tdp();
1240

1241 1242 1243
	if (avic) {
		if (!npt_enabled ||
		    !boot_cpu_has(X86_FEATURE_AVIC) ||
1244
		    !IS_ENABLED(CONFIG_X86_LOCAL_APIC)) {
1245
			avic = false;
1246
		} else {
1247
			pr_info("AVIC enabled\n");
1248 1249 1250

			amd_iommu_register_ga_log_notifier(&avic_ga_log_notifier);
		}
1251
	}
1252

1253 1254
	if (vls) {
		if (!npt_enabled ||
1255
		    !boot_cpu_has(X86_FEATURE_V_VMSAVE_VMLOAD) ||
1256 1257 1258 1259 1260 1261 1262
		    !IS_ENABLED(CONFIG_X86_64)) {
			vls = false;
		} else {
			pr_info("Virtual VMLOAD VMSAVE supported\n");
		}
	}

1263 1264 1265 1266 1267 1268 1269
	if (vgif) {
		if (!boot_cpu_has(X86_FEATURE_VGIF))
			vgif = false;
		else
			pr_info("Virtual GIF supported\n");
	}

A
Avi Kivity 已提交
1270 1271
	return 0;

1272
err:
A
Avi Kivity 已提交
1273 1274 1275 1276 1277 1278 1279
	__free_pages(iopm_pages, IOPM_ALLOC_ORDER);
	iopm_base = 0;
	return r;
}

static __exit void svm_hardware_unsetup(void)
{
1280 1281
	int cpu;

B
Brijesh Singh 已提交
1282 1283 1284
	if (svm_sev_enabled())
		kfree(sev_asid_bitmap);

Z
Zachary Amsden 已提交
1285
	for_each_possible_cpu(cpu)
1286 1287
		svm_cpu_uninit(cpu);

A
Avi Kivity 已提交
1288
	__free_pages(pfn_to_page(iopm_base >> PAGE_SHIFT), IOPM_ALLOC_ORDER);
1289
	iopm_base = 0;
A
Avi Kivity 已提交
1290 1291 1292 1293 1294 1295
}

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
1296
		      SVM_SELECTOR_WRITE_MASK; /* Read/Write Data Segment */
A
Avi Kivity 已提交
1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
	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;
}

1309 1310 1311 1312 1313
static void svm_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u64 g_tsc_offset = 0;

1314
	if (is_guest_mode(vcpu)) {
1315 1316 1317
		g_tsc_offset = svm->vmcb->control.tsc_offset -
			       svm->nested.hsave->control.tsc_offset;
		svm->nested.hsave->control.tsc_offset = offset;
1318 1319 1320 1321
	} else
		trace_kvm_write_tsc_offset(vcpu->vcpu_id,
					   svm->vmcb->control.tsc_offset,
					   offset);
1322 1323

	svm->vmcb->control.tsc_offset = offset + g_tsc_offset;
1324 1325

	mark_dirty(svm->vmcb, VMCB_INTERCEPTS);
1326 1327
}

1328 1329 1330 1331
static void avic_init_vmcb(struct vcpu_svm *svm)
{
	struct vmcb *vmcb = svm->vmcb;
	struct kvm_arch *vm_data = &svm->vcpu.kvm->arch;
1332 1333 1334
	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));
1335 1336 1337 1338 1339 1340 1341 1342

	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;
}

P
Paolo Bonzini 已提交
1343
static void init_vmcb(struct vcpu_svm *svm)
A
Avi Kivity 已提交
1344
{
1345 1346
	struct vmcb_control_area *control = &svm->vmcb->control;
	struct vmcb_save_area *save = &svm->vmcb->save;
A
Avi Kivity 已提交
1347

1348
	svm->vcpu.arch.hflags = 0;
1349

1350 1351 1352 1353 1354 1355
	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);
1356 1357
	if (!kvm_vcpu_apicv_active(&svm->vcpu))
		set_cr_intercept(svm, INTERCEPT_CR8_WRITE);
A
Avi Kivity 已提交
1358

1359
	set_dr_intercepts(svm);
A
Avi Kivity 已提交
1360

1361 1362 1363
	set_exception_intercept(svm, PF_VECTOR);
	set_exception_intercept(svm, UD_VECTOR);
	set_exception_intercept(svm, MC_VECTOR);
1364
	set_exception_intercept(svm, AC_VECTOR);
1365
	set_exception_intercept(svm, DB_VECTOR);
A
Avi Kivity 已提交
1366

1367 1368 1369 1370
	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
1371
	set_intercept(svm, INTERCEPT_RDPMC);
1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388
	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
1389
	set_intercept(svm, INTERCEPT_XSETBV);
B
Brijesh Singh 已提交
1390
	set_intercept(svm, INTERCEPT_RSM);
A
Avi Kivity 已提交
1391

1392 1393 1394 1395 1396
	if (!kvm_mwait_in_guest()) {
		set_intercept(svm, INTERCEPT_MONITOR);
		set_intercept(svm, INTERCEPT_MWAIT);
	}

1397 1398
	control->iopm_base_pa = __sme_set(iopm_base);
	control->msrpm_base_pa = __sme_set(__pa(svm->msrpm));
A
Avi Kivity 已提交
1399 1400 1401 1402 1403 1404 1405 1406 1407
	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;
1408
	save->cs.base = 0xffff0000;
A
Avi Kivity 已提交
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419
	/* 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
1420
	svm_set_efer(&svm->vcpu, 0);
M
Mike Day 已提交
1421
	save->dr6 = 0xffff0ff0;
1422
	kvm_set_rflags(&svm->vcpu, 2);
A
Avi Kivity 已提交
1423
	save->rip = 0x0000fff0;
1424
	svm->vcpu.arch.regs[VCPU_REGS_RIP] = save->rip;
A
Avi Kivity 已提交
1425

J
Joerg Roedel 已提交
1426
	/*
1427
	 * svm_set_cr0() sets PG and WP and clears NW and CD on save->cr0.
1428
	 * It also updates the guest-visible cr0 value.
A
Avi Kivity 已提交
1429
	 */
1430
	svm_set_cr0(&svm->vcpu, X86_CR0_NW | X86_CR0_CD | X86_CR0_ET);
1431
	kvm_mmu_reset_context(&svm->vcpu);
1432

1433
	save->cr4 = X86_CR4_PAE;
A
Avi Kivity 已提交
1434
	/* rdx = ?? */
1435 1436 1437

	if (npt_enabled) {
		/* Setup VMCB for Nested Paging */
1438
		control->nested_ctl |= SVM_NESTED_CTL_NP_ENABLE;
1439
		clr_intercept(svm, INTERCEPT_INVLPG);
1440
		clr_exception_intercept(svm, PF_VECTOR);
1441 1442
		clr_cr_intercept(svm, INTERCEPT_CR3_READ);
		clr_cr_intercept(svm, INTERCEPT_CR3_WRITE);
1443
		save->g_pat = svm->vcpu.arch.pat;
1444 1445 1446
		save->cr3 = 0;
		save->cr4 = 0;
	}
1447
	svm->asid_generation = 0;
1448

1449
	svm->nested.vmcb = 0;
1450 1451
	svm->vcpu.arch.hflags = 0;

1452
	if (boot_cpu_has(X86_FEATURE_PAUSEFILTER)) {
1453
		control->pause_filter_count = 3000;
1454
		set_intercept(svm, INTERCEPT_PAUSE);
1455 1456
	}

1457
	if (kvm_vcpu_apicv_active(&svm->vcpu))
1458 1459
		avic_init_vmcb(svm);

1460 1461 1462 1463 1464 1465 1466 1467 1468 1469
	/*
	 * 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;
	}

1470 1471 1472 1473 1474 1475
	if (vgif) {
		clr_intercept(svm, INTERCEPT_STGI);
		clr_intercept(svm, INTERCEPT_CLGI);
		svm->vmcb->control.int_ctl |= V_GIF_ENABLE_MASK;
	}

1476
	if (sev_guest(svm->vcpu.kvm)) {
B
Brijesh Singh 已提交
1477
		svm->vmcb->control.nested_ctl |= SVM_NESTED_CTL_SEV_ENABLE;
1478 1479
		clr_exception_intercept(svm, UD_VECTOR);
	}
B
Brijesh Singh 已提交
1480

1481 1482
	mark_all_dirty(svm->vmcb);

1483
	enable_gif(svm);
1484 1485 1486

}

1487 1488
static u64 *avic_get_physical_id_entry(struct kvm_vcpu *vcpu,
				       unsigned int index)
1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552
{
	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);
1553 1554 1555
	new_entry = __sme_set((page_to_phys(svm->avic_backing_page) &
			      AVIC_PHYSICAL_ID_ENTRY_BACKING_PAGE_MASK) |
			      AVIC_PHYSICAL_ID_ENTRY_VALID_MASK);
1556 1557 1558 1559 1560 1561 1562
	WRITE_ONCE(*entry, new_entry);

	svm->avic_physical_id_cache = entry;

	return 0;
}

B
Brijesh Singh 已提交
1563 1564
static void __sev_asid_free(int asid)
{
1565 1566
	struct svm_cpu_data *sd;
	int cpu, pos;
B
Brijesh Singh 已提交
1567 1568 1569

	pos = asid - 1;
	clear_bit(pos, sev_asid_bitmap);
1570 1571 1572 1573 1574

	for_each_possible_cpu(cpu) {
		sd = per_cpu(svm_data, cpu);
		sd->sev_vmcbs[pos] = NULL;
	}
B
Brijesh Singh 已提交
1575 1576 1577 1578 1579 1580 1581 1582 1583
}

static void sev_asid_free(struct kvm *kvm)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;

	__sev_asid_free(sev->asid);
}

1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614
static void sev_unbind_asid(struct kvm *kvm, unsigned int handle)
{
	struct sev_data_decommission *decommission;
	struct sev_data_deactivate *data;

	if (!handle)
		return;

	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		return;

	/* deactivate handle */
	data->handle = handle;
	sev_guest_deactivate(data, NULL);

	wbinvd_on_all_cpus();
	sev_guest_df_flush(NULL);
	kfree(data);

	decommission = kzalloc(sizeof(*decommission), GFP_KERNEL);
	if (!decommission)
		return;

	/* decommission handle */
	decommission->handle = handle;
	sev_guest_decommission(decommission, NULL);

	kfree(decommission);
}

1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691
static struct page **sev_pin_memory(struct kvm *kvm, unsigned long uaddr,
				    unsigned long ulen, unsigned long *n,
				    int write)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	unsigned long npages, npinned, size;
	unsigned long locked, lock_limit;
	struct page **pages;
	int first, last;

	/* Calculate number of pages. */
	first = (uaddr & PAGE_MASK) >> PAGE_SHIFT;
	last = ((uaddr + ulen - 1) & PAGE_MASK) >> PAGE_SHIFT;
	npages = (last - first + 1);

	locked = sev->pages_locked + npages;
	lock_limit = rlimit(RLIMIT_MEMLOCK) >> PAGE_SHIFT;
	if (locked > lock_limit && !capable(CAP_IPC_LOCK)) {
		pr_err("SEV: %lu locked pages exceed the lock limit of %lu.\n", locked, lock_limit);
		return NULL;
	}

	/* Avoid using vmalloc for smaller buffers. */
	size = npages * sizeof(struct page *);
	if (size > PAGE_SIZE)
		pages = vmalloc(size);
	else
		pages = kmalloc(size, GFP_KERNEL);

	if (!pages)
		return NULL;

	/* Pin the user virtual address. */
	npinned = get_user_pages_fast(uaddr, npages, write ? FOLL_WRITE : 0, pages);
	if (npinned != npages) {
		pr_err("SEV: Failure locking %lu pages.\n", npages);
		goto err;
	}

	*n = npages;
	sev->pages_locked = locked;

	return pages;

err:
	if (npinned > 0)
		release_pages(pages, npinned);

	kvfree(pages);
	return NULL;
}

static void sev_unpin_memory(struct kvm *kvm, struct page **pages,
			     unsigned long npages)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;

	release_pages(pages, npages);
	kvfree(pages);
	sev->pages_locked -= npages;
}

static void sev_clflush_pages(struct page *pages[], unsigned long npages)
{
	uint8_t *page_virtual;
	unsigned long i;

	if (npages == 0 || pages == NULL)
		return;

	for (i = 0; i < npages; i++) {
		page_virtual = kmap_atomic(pages[i]);
		clflush_cache_range(page_virtual, PAGE_SIZE);
		kunmap_atomic(page_virtual);
	}
}

1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707
static void __unregister_enc_region_locked(struct kvm *kvm,
					   struct enc_region *region)
{
	/*
	 * The guest may change the memory encryption attribute from C=0 -> C=1
	 * or vice versa for this memory range. Lets make sure caches are
	 * flushed to ensure that guest data gets written into memory with
	 * correct C-bit.
	 */
	sev_clflush_pages(region->pages, region->npages);

	sev_unpin_memory(kvm, region->pages, region->npages);
	list_del(&region->list);
	kfree(region);
}

B
Brijesh Singh 已提交
1708 1709
static void sev_vm_destroy(struct kvm *kvm)
{
1710
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
1711 1712
	struct list_head *head = &sev->regions_list;
	struct list_head *pos, *q;
1713

B
Brijesh Singh 已提交
1714 1715 1716
	if (!sev_guest(kvm))
		return;

1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731
	mutex_lock(&kvm->lock);

	/*
	 * if userspace was terminated before unregistering the memory regions
	 * then lets unpin all the registered memory.
	 */
	if (!list_empty(head)) {
		list_for_each_safe(pos, q, head) {
			__unregister_enc_region_locked(kvm,
				list_entry(pos, struct enc_region, list));
		}
	}

	mutex_unlock(&kvm->lock);

1732
	sev_unbind_asid(kvm, sev->handle);
B
Brijesh Singh 已提交
1733 1734 1735
	sev_asid_free(kvm);
}

1736 1737
static void avic_vm_destroy(struct kvm *kvm)
{
1738
	unsigned long flags;
1739 1740
	struct kvm_arch *vm_data = &kvm->arch;

1741 1742 1743
	if (!avic)
		return;

1744 1745 1746 1747
	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);
1748 1749 1750 1751

	spin_lock_irqsave(&svm_vm_data_hash_lock, flags);
	hash_del(&vm_data->hnode);
	spin_unlock_irqrestore(&svm_vm_data_hash_lock, flags);
1752 1753
}

B
Brijesh Singh 已提交
1754 1755 1756 1757 1758 1759
static void svm_vm_destroy(struct kvm *kvm)
{
	avic_vm_destroy(kvm);
	sev_vm_destroy(kvm);
}

1760 1761
static int avic_vm_init(struct kvm *kvm)
{
1762
	unsigned long flags;
1763
	int err = -ENOMEM;
1764 1765 1766
	struct kvm_arch *vm_data = &kvm->arch;
	struct page *p_page;
	struct page *l_page;
1767 1768
	struct kvm_arch *ka;
	u32 vm_id;
1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788

	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));

1789
	spin_lock_irqsave(&svm_vm_data_hash_lock, flags);
1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805
 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;
1806 1807 1808
	hash_add(svm_vm_data_hash, &vm_data->hnode, vm_data->avic_vm_id);
	spin_unlock_irqrestore(&svm_vm_data_hash_lock, flags);

1809 1810 1811 1812 1813
	return 0;

free_avic:
	avic_vm_destroy(kvm);
	return err;
A
Avi Kivity 已提交
1814 1815
}

1816 1817
static inline int
avic_update_iommu_vcpu_affinity(struct kvm_vcpu *vcpu, int cpu, bool r)
1818
{
1819 1820 1821
	int ret = 0;
	unsigned long flags;
	struct amd_svm_iommu_ir *ir;
1822 1823
	struct vcpu_svm *svm = to_svm(vcpu);

1824 1825
	if (!kvm_arch_has_assigned_device(vcpu->kvm))
		return 0;
1826

1827 1828 1829 1830 1831
	/*
	 * 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);
1832

1833 1834
	if (list_empty(&svm->ir_list))
		goto out;
1835

1836 1837 1838 1839 1840 1841 1842 1843
	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;
1844 1845 1846 1847 1848 1849
}

static void avic_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
	u64 entry;
	/* ID = 0xff (broadcast), ID > 0xff (reserved) */
1850
	int h_physical_id = kvm_cpu_get_apicid(cpu);
1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869
	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);
1870 1871
	avic_update_iommu_vcpu_affinity(vcpu, h_physical_id,
					svm->avic_is_running);
1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882
}

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));
1883 1884 1885
	if (entry & AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK)
		avic_update_iommu_vcpu_affinity(vcpu, -1, 0);

1886 1887
	entry &= ~AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK;
	WRITE_ONCE(*(svm->avic_physical_id_cache), entry);
A
Avi Kivity 已提交
1888 1889
}

1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903
/**
 * 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);
}

1904
static void svm_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)
1905 1906
{
	struct vcpu_svm *svm = to_svm(vcpu);
1907 1908
	u32 dummy;
	u32 eax = 1;
1909

1910 1911
	svm->spec_ctrl = 0;

1912 1913 1914 1915 1916 1917
	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
1918
	init_vmcb(svm);
A
Avi Kivity 已提交
1919

1920
	kvm_cpuid(vcpu, &eax, &dummy, &dummy, &dummy, true);
1921
	kvm_register_write(vcpu, VCPU_REGS_RDX, eax);
1922 1923 1924

	if (kvm_vcpu_apicv_active(vcpu) && !init_event)
		avic_update_vapic_bar(svm, APIC_DEFAULT_PHYS_BASE);
1925 1926
}

1927 1928 1929 1930
static int avic_init_vcpu(struct vcpu_svm *svm)
{
	int ret;

1931
	if (!kvm_vcpu_apicv_active(&svm->vcpu))
1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943
		return 0;

	ret = avic_init_backing_page(&svm->vcpu);
	if (ret)
		return ret;

	INIT_LIST_HEAD(&svm->ir_list);
	spin_lock_init(&svm->ir_list_lock);

	return ret;
}

R
Rusty Russell 已提交
1944
static struct kvm_vcpu *svm_create_vcpu(struct kvm *kvm, unsigned int id)
A
Avi Kivity 已提交
1945
{
1946
	struct vcpu_svm *svm;
A
Avi Kivity 已提交
1947
	struct page *page;
1948
	struct page *msrpm_pages;
A
Alexander Graf 已提交
1949
	struct page *hsave_page;
A
Alexander Graf 已提交
1950
	struct page *nested_msrpm_pages;
R
Rusty Russell 已提交
1951
	int err;
A
Avi Kivity 已提交
1952

1953
	svm = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL);
R
Rusty Russell 已提交
1954 1955 1956 1957 1958 1959 1960 1961 1962
	if (!svm) {
		err = -ENOMEM;
		goto out;
	}

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

1963
	err = -ENOMEM;
A
Avi Kivity 已提交
1964
	page = alloc_page(GFP_KERNEL);
1965
	if (!page)
R
Rusty Russell 已提交
1966
		goto uninit;
A
Avi Kivity 已提交
1967

1968 1969
	msrpm_pages = alloc_pages(GFP_KERNEL, MSRPM_ALLOC_ORDER);
	if (!msrpm_pages)
1970
		goto free_page1;
A
Alexander Graf 已提交
1971 1972 1973

	nested_msrpm_pages = alloc_pages(GFP_KERNEL, MSRPM_ALLOC_ORDER);
	if (!nested_msrpm_pages)
1974
		goto free_page2;
1975

A
Alexander Graf 已提交
1976 1977
	hsave_page = alloc_page(GFP_KERNEL);
	if (!hsave_page)
1978 1979
		goto free_page3;

1980 1981 1982
	err = avic_init_vcpu(svm);
	if (err)
		goto free_page4;
1983

1984 1985 1986 1987 1988
	/* 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;

1989
	svm->nested.hsave = page_address(hsave_page);
A
Alexander Graf 已提交
1990

1991 1992 1993
	svm->msrpm = page_address(msrpm_pages);
	svm_vcpu_init_msrpm(svm->msrpm);

1994
	svm->nested.msrpm = page_address(nested_msrpm_pages);
1995
	svm_vcpu_init_msrpm(svm->nested.msrpm);
A
Alexander Graf 已提交
1996

1997 1998
	svm->vmcb = page_address(page);
	clear_page(svm->vmcb);
1999
	svm->vmcb_pa = __sme_set(page_to_pfn(page) << PAGE_SHIFT);
2000
	svm->asid_generation = 0;
P
Paolo Bonzini 已提交
2001
	init_vmcb(svm);
A
Avi Kivity 已提交
2002

2003 2004
	svm_init_osvw(&svm->vcpu);

R
Rusty Russell 已提交
2005
	return &svm->vcpu;
2006

2007 2008
free_page4:
	__free_page(hsave_page);
2009 2010 2011 2012 2013 2014
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
2015 2016 2017
uninit:
	kvm_vcpu_uninit(&svm->vcpu);
free_svm:
2018
	kmem_cache_free(kvm_vcpu_cache, svm);
R
Rusty Russell 已提交
2019 2020
out:
	return ERR_PTR(err);
A
Avi Kivity 已提交
2021 2022 2023 2024
}

static void svm_free_vcpu(struct kvm_vcpu *vcpu)
{
2025 2026
	struct vcpu_svm *svm = to_svm(vcpu);

2027
	__free_page(pfn_to_page(__sme_clr(svm->vmcb_pa) >> PAGE_SHIFT));
2028
	__free_pages(virt_to_page(svm->msrpm), MSRPM_ALLOC_ORDER);
2029 2030
	__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
2031
	kvm_vcpu_uninit(vcpu);
2032
	kmem_cache_free(kvm_vcpu_cache, svm);
A
Ashok Raj 已提交
2033 2034 2035 2036 2037
	/*
	 * The vmcb page can be recycled, causing a false negative in
	 * svm_vcpu_load(). So do a full IBPB now.
	 */
	indirect_branch_prediction_barrier();
A
Avi Kivity 已提交
2038 2039
}

2040
static void svm_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
A
Avi Kivity 已提交
2041
{
2042
	struct vcpu_svm *svm = to_svm(vcpu);
A
Ashok Raj 已提交
2043
	struct svm_cpu_data *sd = per_cpu(svm_data, cpu);
2044
	int i;
2045 2046

	if (unlikely(cpu != vcpu->cpu)) {
2047
		svm->asid_generation = 0;
2048
		mark_all_dirty(svm->vmcb);
2049
	}
2050

2051 2052 2053
#ifdef CONFIG_X86_64
	rdmsrl(MSR_GS_BASE, to_svm(vcpu)->host.gs_base);
#endif
2054 2055 2056 2057
	savesegment(fs, svm->host.fs);
	savesegment(gs, svm->host.gs);
	svm->host.ldt = kvm_read_ldt();

2058
	for (i = 0; i < NR_HOST_SAVE_USER_MSRS; i++)
2059
		rdmsrl(host_save_user_msrs[i], svm->host_user_msrs[i]);
2060

2061 2062 2063 2064 2065 2066
	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);
		}
2067
	}
P
Paolo Bonzini 已提交
2068 2069 2070
	/* This assumes that the kernel never uses MSR_TSC_AUX */
	if (static_cpu_has(X86_FEATURE_RDTSCP))
		wrmsrl(MSR_TSC_AUX, svm->tsc_aux);
2071

A
Ashok Raj 已提交
2072 2073 2074 2075
	if (sd->current_vmcb != svm->vmcb) {
		sd->current_vmcb = svm->vmcb;
		indirect_branch_prediction_barrier();
	}
2076
	avic_vcpu_load(vcpu, cpu);
A
Avi Kivity 已提交
2077 2078 2079 2080
}

static void svm_vcpu_put(struct kvm_vcpu *vcpu)
{
2081
	struct vcpu_svm *svm = to_svm(vcpu);
2082 2083
	int i;

2084 2085
	avic_vcpu_put(vcpu);

2086
	++vcpu->stat.host_state_reload;
2087 2088 2089
	kvm_load_ldt(svm->host.ldt);
#ifdef CONFIG_X86_64
	loadsegment(fs, svm->host.fs);
2090
	wrmsrl(MSR_KERNEL_GS_BASE, current->thread.gsbase);
2091
	load_gs_index(svm->host.gs);
2092
#else
2093
#ifdef CONFIG_X86_32_LAZY_GS
2094
	loadsegment(gs, svm->host.gs);
2095
#endif
2096
#endif
2097
	for (i = 0; i < NR_HOST_SAVE_USER_MSRS; i++)
2098
		wrmsrl(host_save_user_msrs[i], svm->host_user_msrs[i]);
A
Avi Kivity 已提交
2099 2100
}

2101 2102 2103 2104 2105 2106 2107 2108 2109 2110
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
2111 2112
static unsigned long svm_get_rflags(struct kvm_vcpu *vcpu)
{
2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123
	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
2124 2125 2126 2127
}

static void svm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags)
{
2128 2129 2130
	if (to_svm(vcpu)->nmi_singlestep)
		rflags |= (X86_EFLAGS_TF | X86_EFLAGS_RF);

P
Paolo Bonzini 已提交
2131
       /*
A
Andrea Gelmini 已提交
2132
        * Any change of EFLAGS.VM is accompanied by a reload of SS
P
Paolo Bonzini 已提交
2133 2134 2135
        * (caused by either a task switch or an inter-privilege IRET),
        * so we do not need to update the CPL here.
        */
2136
	to_svm(vcpu)->vmcb->save.rflags = rflags;
A
Avi Kivity 已提交
2137 2138
}

A
Avi Kivity 已提交
2139 2140 2141 2142 2143
static void svm_cache_reg(struct kvm_vcpu *vcpu, enum kvm_reg reg)
{
	switch (reg) {
	case VCPU_EXREG_PDPTR:
		BUG_ON(!npt_enabled);
2144
		load_pdptrs(vcpu, vcpu->arch.walk_mmu, kvm_read_cr3(vcpu));
A
Avi Kivity 已提交
2145 2146 2147 2148 2149 2150
		break;
	default:
		BUG();
	}
}

2151 2152
static void svm_set_vintr(struct vcpu_svm *svm)
{
2153
	set_intercept(svm, INTERCEPT_VINTR);
2154 2155 2156 2157
}

static void svm_clear_vintr(struct vcpu_svm *svm)
{
2158
	clr_intercept(svm, INTERCEPT_VINTR);
2159 2160
}

A
Avi Kivity 已提交
2161 2162
static struct vmcb_seg *svm_seg(struct kvm_vcpu *vcpu, int seg)
{
2163
	struct vmcb_save_area *save = &to_svm(vcpu)->vmcb->save;
A
Avi Kivity 已提交
2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175

	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
2176
	return NULL;
A
Avi Kivity 已提交
2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200
}

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;
2201 2202 2203 2204 2205 2206 2207 2208 2209 2210

	/*
	 * 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;
2211

J
Joerg Roedel 已提交
2212 2213
	/*
	 * AMD's VMCB does not have an explicit unusable field, so emulate it
2214 2215
	 * for cross vendor migration purposes by "not present"
	 */
2216
	var->unusable = !var->present;
2217

2218 2219 2220 2221 2222 2223
	switch (seg) {
	case VCPU_SREG_TR:
		/*
		 * Work around a bug where the busy flag in the tr selector
		 * isn't exposed
		 */
2224
		var->type |= 0x2;
2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239
		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;
2240
	case VCPU_SREG_SS:
J
Joerg Roedel 已提交
2241 2242
		/*
		 * On AMD CPUs sometimes the DB bit in the segment
2243 2244 2245 2246 2247 2248
		 * 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;
2249
		/* This is symmetric with svm_set_segment() */
J
Jan Kiszka 已提交
2250
		var->dpl = to_svm(vcpu)->vmcb->save.cpl;
2251
		break;
2252
	}
A
Avi Kivity 已提交
2253 2254
}

2255 2256 2257 2258 2259 2260 2261
static int svm_get_cpl(struct kvm_vcpu *vcpu)
{
	struct vmcb_save_area *save = &to_svm(vcpu)->vmcb->save;

	return save->cpl;
}

2262
static void svm_get_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
2263
{
2264 2265
	struct vcpu_svm *svm = to_svm(vcpu);

2266 2267
	dt->size = svm->vmcb->save.idtr.limit;
	dt->address = svm->vmcb->save.idtr.base;
A
Avi Kivity 已提交
2268 2269
}

2270
static void svm_set_idt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
2271
{
2272 2273
	struct vcpu_svm *svm = to_svm(vcpu);

2274 2275
	svm->vmcb->save.idtr.limit = dt->size;
	svm->vmcb->save.idtr.base = dt->address ;
2276
	mark_dirty(svm->vmcb, VMCB_DT);
A
Avi Kivity 已提交
2277 2278
}

2279
static void svm_get_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
2280
{
2281 2282
	struct vcpu_svm *svm = to_svm(vcpu);

2283 2284
	dt->size = svm->vmcb->save.gdtr.limit;
	dt->address = svm->vmcb->save.gdtr.base;
A
Avi Kivity 已提交
2285 2286
}

2287
static void svm_set_gdt(struct kvm_vcpu *vcpu, struct desc_ptr *dt)
A
Avi Kivity 已提交
2288
{
2289 2290
	struct vcpu_svm *svm = to_svm(vcpu);

2291 2292
	svm->vmcb->save.gdtr.limit = dt->size;
	svm->vmcb->save.gdtr.base = dt->address ;
2293
	mark_dirty(svm->vmcb, VMCB_DT);
A
Avi Kivity 已提交
2294 2295
}

2296 2297 2298 2299
static void svm_decache_cr0_guest_bits(struct kvm_vcpu *vcpu)
{
}

2300 2301 2302 2303
static void svm_decache_cr3(struct kvm_vcpu *vcpu)
{
}

2304
static void svm_decache_cr4_guest_bits(struct kvm_vcpu *vcpu)
2305 2306 2307
{
}

A
Avi Kivity 已提交
2308 2309 2310 2311 2312
static void update_cr0_intercept(struct vcpu_svm *svm)
{
	ulong gcr0 = svm->vcpu.arch.cr0;
	u64 *hcr0 = &svm->vmcb->save.cr0;

2313 2314
	*hcr0 = (*hcr0 & ~SVM_CR0_SELECTIVE_MASK)
		| (gcr0 & SVM_CR0_SELECTIVE_MASK);
A
Avi Kivity 已提交
2315

2316
	mark_dirty(svm->vmcb, VMCB_CR);
A
Avi Kivity 已提交
2317

2318
	if (gcr0 == *hcr0) {
2319 2320
		clr_cr_intercept(svm, INTERCEPT_CR0_READ);
		clr_cr_intercept(svm, INTERCEPT_CR0_WRITE);
A
Avi Kivity 已提交
2321
	} else {
2322 2323
		set_cr_intercept(svm, INTERCEPT_CR0_READ);
		set_cr_intercept(svm, INTERCEPT_CR0_WRITE);
A
Avi Kivity 已提交
2324 2325 2326
	}
}

A
Avi Kivity 已提交
2327 2328
static void svm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
{
2329 2330
	struct vcpu_svm *svm = to_svm(vcpu);

2331
#ifdef CONFIG_X86_64
2332
	if (vcpu->arch.efer & EFER_LME) {
2333
		if (!is_paging(vcpu) && (cr0 & X86_CR0_PG)) {
2334
			vcpu->arch.efer |= EFER_LMA;
2335
			svm->vmcb->save.efer |= EFER_LMA | EFER_LME;
A
Avi Kivity 已提交
2336 2337
		}

M
Mike Day 已提交
2338
		if (is_paging(vcpu) && !(cr0 & X86_CR0_PG)) {
2339
			vcpu->arch.efer &= ~EFER_LMA;
2340
			svm->vmcb->save.efer &= ~(EFER_LMA | EFER_LME);
A
Avi Kivity 已提交
2341 2342 2343
		}
	}
#endif
2344
	vcpu->arch.cr0 = cr0;
2345 2346 2347

	if (!npt_enabled)
		cr0 |= X86_CR0_PG | X86_CR0_WP;
2348

2349 2350 2351 2352 2353 2354 2355
	/*
	 * 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);
2356
	svm->vmcb->save.cr0 = cr0;
2357
	mark_dirty(svm->vmcb, VMCB_CR);
A
Avi Kivity 已提交
2358
	update_cr0_intercept(svm);
A
Avi Kivity 已提交
2359 2360
}

2361
static int svm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
A
Avi Kivity 已提交
2362
{
2363
	unsigned long host_cr4_mce = cr4_read_shadow() & X86_CR4_MCE;
2364 2365
	unsigned long old_cr4 = to_svm(vcpu)->vmcb->save.cr4;

2366 2367 2368
	if (cr4 & X86_CR4_VMXE)
		return 1;

2369
	if (npt_enabled && ((old_cr4 ^ cr4) & X86_CR4_PGE))
2370
		svm_flush_tlb(vcpu, true);
2371

2372 2373 2374
	vcpu->arch.cr4 = cr4;
	if (!npt_enabled)
		cr4 |= X86_CR4_PAE;
2375
	cr4 |= host_cr4_mce;
2376
	to_svm(vcpu)->vmcb->save.cr4 = cr4;
2377
	mark_dirty(to_svm(vcpu)->vmcb, VMCB_CR);
2378
	return 0;
A
Avi Kivity 已提交
2379 2380 2381 2382 2383
}

static void svm_set_segment(struct kvm_vcpu *vcpu,
			    struct kvm_segment *var, int seg)
{
2384
	struct vcpu_svm *svm = to_svm(vcpu);
A
Avi Kivity 已提交
2385 2386 2387 2388 2389
	struct vmcb_seg *s = svm_seg(vcpu, seg);

	s->base = var->base;
	s->limit = var->limit;
	s->selector = var->selector;
2390 2391 2392 2393 2394 2395 2396 2397
	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
2398 2399 2400 2401 2402 2403 2404 2405

	/*
	 * 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)
2406 2407
		/* This is symmetric with svm_get_segment() */
		svm->vmcb->save.cpl = (var->dpl & 3);
A
Avi Kivity 已提交
2408

2409
	mark_dirty(svm->vmcb, VMCB_SEG);
A
Avi Kivity 已提交
2410 2411
}

2412
static void update_bp_intercept(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
2413
{
J
Jan Kiszka 已提交
2414 2415
	struct vcpu_svm *svm = to_svm(vcpu);

2416
	clr_exception_intercept(svm, BP_VECTOR);
2417

J
Jan Kiszka 已提交
2418 2419
	if (vcpu->guest_debug & KVM_GUESTDBG_ENABLE) {
		if (vcpu->guest_debug & KVM_GUESTDBG_USE_SW_BP)
2420
			set_exception_intercept(svm, BP_VECTOR);
J
Jan Kiszka 已提交
2421 2422
	} else
		vcpu->guest_debug = 0;
2423 2424
}

2425
static void new_asid(struct vcpu_svm *svm, struct svm_cpu_data *sd)
A
Avi Kivity 已提交
2426
{
2427 2428
	if (sd->next_asid > sd->max_asid) {
		++sd->asid_generation;
2429
		sd->next_asid = sd->min_asid;
2430
		svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ALL_ASID;
A
Avi Kivity 已提交
2431 2432
	}

2433 2434
	svm->asid_generation = sd->asid_generation;
	svm->vmcb->control.asid = sd->next_asid++;
2435 2436

	mark_dirty(svm->vmcb, VMCB_ASID);
A
Avi Kivity 已提交
2437 2438
}

J
Jan Kiszka 已提交
2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451
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);
}

2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466
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);
}

2467
static void svm_set_dr7(struct kvm_vcpu *vcpu, unsigned long value)
A
Avi Kivity 已提交
2468
{
2469 2470
	struct vcpu_svm *svm = to_svm(vcpu);

2471
	svm->vmcb->save.dr7 = value;
2472
	mark_dirty(svm->vmcb, VMCB_DR);
A
Avi Kivity 已提交
2473 2474
}

A
Avi Kivity 已提交
2475
static int pf_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2476
{
2477
	u64 fault_address = __sme_clr(svm->vmcb->control.exit_info_2);
2478
	u64 error_code = svm->vmcb->control.exit_info_1;
A
Avi Kivity 已提交
2479

2480
	return kvm_handle_page_fault(&svm->vcpu, error_code, fault_address,
2481 2482
			static_cpu_has(X86_FEATURE_DECODEASSISTS) ?
			svm->vmcb->control.insn_bytes : NULL,
2483 2484 2485 2486 2487
			svm->vmcb->control.insn_len);
}

static int npf_interception(struct vcpu_svm *svm)
{
2488
	u64 fault_address = __sme_clr(svm->vmcb->control.exit_info_2);
2489 2490 2491 2492
	u64 error_code = svm->vmcb->control.exit_info_1;

	trace_kvm_page_fault(fault_address, error_code);
	return kvm_mmu_page_fault(&svm->vcpu, fault_address, error_code,
2493 2494
			static_cpu_has(X86_FEATURE_DECODEASSISTS) ?
			svm->vmcb->control.insn_bytes : NULL,
2495
			svm->vmcb->control.insn_len);
A
Avi Kivity 已提交
2496 2497
}

A
Avi Kivity 已提交
2498
static int db_interception(struct vcpu_svm *svm)
J
Jan Kiszka 已提交
2499
{
A
Avi Kivity 已提交
2500 2501
	struct kvm_run *kvm_run = svm->vcpu.run;

J
Jan Kiszka 已提交
2502
	if (!(svm->vcpu.guest_debug &
2503
	      (KVM_GUESTDBG_SINGLESTEP | KVM_GUESTDBG_USE_HW_BP)) &&
J
Jan Kiszka 已提交
2504
		!svm->nmi_singlestep) {
J
Jan Kiszka 已提交
2505 2506 2507
		kvm_queue_exception(&svm->vcpu, DB_VECTOR);
		return 1;
	}
2508

J
Jan Kiszka 已提交
2509
	if (svm->nmi_singlestep) {
2510
		disable_nmi_singlestep(svm);
2511 2512 2513
	}

	if (svm->vcpu.guest_debug &
J
Joerg Roedel 已提交
2514
	    (KVM_GUESTDBG_SINGLESTEP | KVM_GUESTDBG_USE_HW_BP)) {
2515 2516 2517 2518 2519 2520 2521 2522
		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
2523 2524
}

A
Avi Kivity 已提交
2525
static int bp_interception(struct vcpu_svm *svm)
J
Jan Kiszka 已提交
2526
{
A
Avi Kivity 已提交
2527 2528
	struct kvm_run *kvm_run = svm->vcpu.run;

J
Jan Kiszka 已提交
2529 2530 2531 2532 2533 2534
	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
2535
static int ud_interception(struct vcpu_svm *svm)
2536 2537 2538
{
	int er;

2539
	er = emulate_instruction(&svm->vcpu, EMULTYPE_TRAP_UD);
2540 2541
	if (er == EMULATE_USER_EXIT)
		return 0;
2542
	if (er != EMULATE_DONE)
2543
		kvm_queue_exception(&svm->vcpu, UD_VECTOR);
2544 2545 2546
	return 1;
}

2547 2548 2549 2550 2551 2552
static int ac_interception(struct vcpu_svm *svm)
{
	kvm_queue_exception_e(&svm->vcpu, AC_VECTOR, 0);
	return 1;
}

2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591
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;
}

2592
static void svm_handle_mce(struct vcpu_svm *svm)
2593
{
2594 2595 2596 2597 2598 2599 2600
	if (is_erratum_383()) {
		/*
		 * Erratum 383 triggered. Guest state is corrupt so kill the
		 * guest.
		 */
		pr_err("KVM: Guest triggered AMD Erratum 383\n");

2601
		kvm_make_request(KVM_REQ_TRIPLE_FAULT, &svm->vcpu);
2602 2603 2604 2605

		return;
	}

2606 2607 2608 2609 2610 2611 2612 2613
	/*
	 * 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 */

2614 2615 2616 2617 2618
	return;
}

static int mc_interception(struct vcpu_svm *svm)
{
2619 2620 2621
	return 1;
}

A
Avi Kivity 已提交
2622
static int shutdown_interception(struct vcpu_svm *svm)
2623
{
A
Avi Kivity 已提交
2624 2625
	struct kvm_run *kvm_run = svm->vcpu.run;

2626 2627 2628 2629
	/*
	 * VMCB is undefined after a SHUTDOWN intercept
	 * so reinitialize it.
	 */
2630
	clear_page(svm->vmcb);
P
Paolo Bonzini 已提交
2631
	init_vmcb(svm);
2632 2633 2634 2635 2636

	kvm_run->exit_reason = KVM_EXIT_SHUTDOWN;
	return 0;
}

A
Avi Kivity 已提交
2637
static int io_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2638
{
2639
	struct kvm_vcpu *vcpu = &svm->vcpu;
M
Mike Day 已提交
2640
	u32 io_info = svm->vmcb->control.exit_info_1; /* address size bug? */
2641
	int size, in, string, ret;
2642
	unsigned port;
A
Avi Kivity 已提交
2643

R
Rusty Russell 已提交
2644
	++svm->vcpu.stat.io_exits;
2645
	string = (io_info & SVM_IOIO_STR_MASK) != 0;
2646
	in = (io_info & SVM_IOIO_TYPE_MASK) != 0;
2647
	if (string)
2648
		return emulate_instruction(vcpu, 0) == EMULATE_DONE;
2649

2650 2651
	port = io_info >> 16;
	size = (io_info & SVM_IOIO_SIZE_MASK) >> SVM_IOIO_SIZE_SHIFT;
2652
	svm->next_rip = svm->vmcb->control.exit_info_2;
2653
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
2654

2655 2656 2657 2658 2659 2660 2661 2662
	/*
	 * 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
2663 2664
}

A
Avi Kivity 已提交
2665
static int nmi_interception(struct vcpu_svm *svm)
2666 2667 2668 2669
{
	return 1;
}

A
Avi Kivity 已提交
2670
static int intr_interception(struct vcpu_svm *svm)
2671 2672 2673 2674 2675
{
	++svm->vcpu.stat.irq_exits;
	return 1;
}

A
Avi Kivity 已提交
2676
static int nop_on_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2677 2678 2679 2680
{
	return 1;
}

A
Avi Kivity 已提交
2681
static int halt_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
2682
{
2683
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 1;
R
Rusty Russell 已提交
2684
	return kvm_emulate_halt(&svm->vcpu);
A
Avi Kivity 已提交
2685 2686
}

A
Avi Kivity 已提交
2687
static int vmmcall_interception(struct vcpu_svm *svm)
2688
{
2689
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
2690
	return kvm_emulate_hypercall(&svm->vcpu);
2691 2692
}

2693 2694 2695 2696 2697 2698 2699
static unsigned long nested_svm_get_tdp_cr3(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	return svm->nested.nested_cr3;
}

2700 2701 2702 2703 2704 2705 2706
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;

2707
	ret = kvm_vcpu_read_guest_page(vcpu, gpa_to_gfn(__sme_clr(cr3)), &pdpte,
2708
				       offset_in_page(cr3) + index * 8, 8);
2709 2710 2711 2712 2713
	if (ret)
		return 0;
	return pdpte;
}

2714 2715 2716 2717 2718
static void nested_svm_set_tdp_cr3(struct kvm_vcpu *vcpu,
				   unsigned long root)
{
	struct vcpu_svm *svm = to_svm(vcpu);

2719
	svm->vmcb->control.nested_cr3 = __sme_set(root);
2720
	mark_dirty(svm->vmcb, VMCB_NPT);
2721
	svm_flush_tlb(vcpu, true);
2722 2723
}

2724 2725
static void nested_svm_inject_npf_exit(struct kvm_vcpu *vcpu,
				       struct x86_exception *fault)
2726 2727 2728
{
	struct vcpu_svm *svm = to_svm(vcpu);

2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748
	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;
2749 2750 2751 2752

	nested_svm_vmexit(svm);
}

2753
static void nested_svm_init_mmu_context(struct kvm_vcpu *vcpu)
2754
{
2755 2756
	WARN_ON(mmu_is_nested(vcpu));
	kvm_init_shadow_mmu(vcpu);
2757 2758
	vcpu->arch.mmu.set_cr3           = nested_svm_set_tdp_cr3;
	vcpu->arch.mmu.get_cr3           = nested_svm_get_tdp_cr3;
2759
	vcpu->arch.mmu.get_pdptr         = nested_svm_get_tdp_pdptr;
2760
	vcpu->arch.mmu.inject_page_fault = nested_svm_inject_npf_exit;
2761
	vcpu->arch.mmu.shadow_root_level = get_npt_level(vcpu);
2762
	reset_shadow_zero_bits_mask(vcpu, &vcpu->arch.mmu);
2763 2764 2765 2766 2767 2768 2769 2770
	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;
}

2771 2772
static int nested_svm_check_permissions(struct vcpu_svm *svm)
{
2773 2774
	if (!(svm->vcpu.arch.efer & EFER_SVME) ||
	    !is_paging(&svm->vcpu)) {
2775 2776 2777 2778 2779 2780 2781 2782 2783
		kvm_queue_exception(&svm->vcpu, UD_VECTOR);
		return 1;
	}

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

2784
	return 0;
2785 2786
}

2787 2788 2789
static int nested_svm_check_exception(struct vcpu_svm *svm, unsigned nr,
				      bool has_error_code, u32 error_code)
{
2790 2791
	int vmexit;

2792
	if (!is_guest_mode(&svm->vcpu))
2793
		return 0;
2794

2795 2796 2797 2798
	vmexit = nested_svm_intercept(svm);
	if (vmexit != NESTED_EXIT_DONE)
		return 0;

2799 2800 2801
	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;
2802 2803 2804 2805 2806 2807 2808 2809 2810 2811

	/*
	 * 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.
	 */
2812 2813 2814 2815
	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;
2816

2817
	svm->nested.exit_required = true;
2818
	return vmexit;
2819 2820
}

2821 2822
/* This function returns true if it is save to enable the irq window */
static inline bool nested_svm_intr(struct vcpu_svm *svm)
2823
{
2824
	if (!is_guest_mode(&svm->vcpu))
2825
		return true;
2826

2827
	if (!(svm->vcpu.arch.hflags & HF_VINTR_MASK))
2828
		return true;
2829

2830
	if (!(svm->vcpu.arch.hflags & HF_HIF_MASK))
2831
		return false;
2832

2833 2834 2835 2836 2837 2838 2839 2840
	/*
	 * 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;

2841 2842 2843
	svm->vmcb->control.exit_code   = SVM_EXIT_INTR;
	svm->vmcb->control.exit_info_1 = 0;
	svm->vmcb->control.exit_info_2 = 0;
2844

2845 2846 2847
	if (svm->nested.intercept & 1ULL) {
		/*
		 * The #vmexit can't be emulated here directly because this
G
Guo Chao 已提交
2848
		 * code path runs with irqs and preemption disabled. A
2849 2850 2851 2852
		 * #vmexit emulation might sleep. Only signal request for
		 * the #vmexit here.
		 */
		svm->nested.exit_required = true;
2853
		trace_kvm_nested_intr_vmexit(svm->vmcb->save.rip);
2854
		return false;
2855 2856
	}

2857
	return true;
2858 2859
}

2860 2861 2862
/* This function returns true if it is save to enable the nmi window */
static inline bool nested_svm_nmi(struct vcpu_svm *svm)
{
2863
	if (!is_guest_mode(&svm->vcpu))
2864 2865 2866 2867 2868 2869 2870 2871 2872
		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;
2873 2874
}

2875
static void *nested_svm_map(struct vcpu_svm *svm, u64 gpa, struct page **_page)
2876 2877 2878
{
	struct page *page;

2879 2880
	might_sleep();

2881
	page = kvm_vcpu_gfn_to_page(&svm->vcpu, gpa >> PAGE_SHIFT);
2882 2883 2884
	if (is_error_page(page))
		goto error;

2885 2886 2887
	*_page = page;

	return kmap(page);
2888 2889 2890 2891 2892 2893 2894

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

	return NULL;
}

2895
static void nested_svm_unmap(struct page *page)
2896
{
2897
	kunmap(page);
2898 2899 2900
	kvm_release_page_dirty(page);
}

2901 2902
static int nested_svm_intercept_ioio(struct vcpu_svm *svm)
{
2903 2904 2905
	unsigned port, size, iopm_len;
	u16 val, mask;
	u8 start_bit;
2906
	u64 gpa;
2907

2908 2909
	if (!(svm->nested.intercept & (1ULL << INTERCEPT_IOIO_PROT)))
		return NESTED_EXIT_HOST;
2910

2911
	port = svm->vmcb->control.exit_info_1 >> 16;
2912 2913
	size = (svm->vmcb->control.exit_info_1 & SVM_IOIO_SIZE_MASK) >>
		SVM_IOIO_SIZE_SHIFT;
2914
	gpa  = svm->nested.vmcb_iopm + (port / 8);
2915 2916 2917 2918
	start_bit = port % 8;
	iopm_len = (start_bit + size > 8) ? 2 : 1;
	mask = (0xf >> (4 - size)) << start_bit;
	val = 0;
2919

2920
	if (kvm_vcpu_read_guest(&svm->vcpu, gpa, &val, iopm_len))
2921
		return NESTED_EXIT_DONE;
2922

2923
	return (val & mask) ? NESTED_EXIT_DONE : NESTED_EXIT_HOST;
2924 2925
}

2926
static int nested_svm_exit_handled_msr(struct vcpu_svm *svm)
2927
{
2928 2929
	u32 offset, msr, value;
	int write, mask;
2930

2931
	if (!(svm->nested.intercept & (1ULL << INTERCEPT_MSR_PROT)))
2932
		return NESTED_EXIT_HOST;
2933

2934 2935 2936 2937
	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);
2938

2939 2940
	if (offset == MSR_INVALID)
		return NESTED_EXIT_DONE;
2941

2942 2943
	/* Offset is in 32 bit units but need in 8 bit units */
	offset *= 4;
2944

2945
	if (kvm_vcpu_read_guest(&svm->vcpu, svm->nested.vmcb_msrpm + offset, &value, 4))
2946
		return NESTED_EXIT_DONE;
2947

2948
	return (value & mask) ? NESTED_EXIT_DONE : NESTED_EXIT_HOST;
2949 2950
}

2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975
/* 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;
}

2976
static int nested_svm_exit_special(struct vcpu_svm *svm)
2977 2978
{
	u32 exit_code = svm->vmcb->control.exit_code;
2979

2980 2981 2982
	switch (exit_code) {
	case SVM_EXIT_INTR:
	case SVM_EXIT_NMI:
2983
	case SVM_EXIT_EXCP_BASE + MC_VECTOR:
2984 2985
		return NESTED_EXIT_HOST;
	case SVM_EXIT_NPF:
J
Joerg Roedel 已提交
2986
		/* For now we are always handling NPFs when using them */
2987 2988 2989 2990
		if (npt_enabled)
			return NESTED_EXIT_HOST;
		break;
	case SVM_EXIT_EXCP_BASE + PF_VECTOR:
G
Gleb Natapov 已提交
2991
		/* When we're shadowing, trap PFs, but not async PF */
2992
		if (!npt_enabled && svm->vcpu.arch.apf.host_apf_reason == 0)
2993 2994 2995 2996
			return NESTED_EXIT_HOST;
		break;
	default:
		break;
2997 2998
	}

2999 3000 3001 3002 3003 3004
	return NESTED_EXIT_CONTINUE;
}

/*
 * If this function returns true, this #vmexit was already handled
 */
3005
static int nested_svm_intercept(struct vcpu_svm *svm)
3006 3007 3008 3009
{
	u32 exit_code = svm->vmcb->control.exit_code;
	int vmexit = NESTED_EXIT_HOST;

3010
	switch (exit_code) {
3011
	case SVM_EXIT_MSR:
3012
		vmexit = nested_svm_exit_handled_msr(svm);
3013
		break;
3014 3015 3016
	case SVM_EXIT_IOIO:
		vmexit = nested_svm_intercept_ioio(svm);
		break;
3017 3018 3019
	case SVM_EXIT_READ_CR0 ... SVM_EXIT_WRITE_CR8: {
		u32 bit = 1U << (exit_code - SVM_EXIT_READ_CR0);
		if (svm->nested.intercept_cr & bit)
3020
			vmexit = NESTED_EXIT_DONE;
3021 3022
		break;
	}
3023 3024 3025
	case SVM_EXIT_READ_DR0 ... SVM_EXIT_WRITE_DR7: {
		u32 bit = 1U << (exit_code - SVM_EXIT_READ_DR0);
		if (svm->nested.intercept_dr & bit)
3026
			vmexit = NESTED_EXIT_DONE;
3027 3028 3029 3030
		break;
	}
	case SVM_EXIT_EXCP_BASE ... SVM_EXIT_EXCP_BASE + 0x1f: {
		u32 excp_bits = 1 << (exit_code - SVM_EXIT_EXCP_BASE);
3031 3032 3033 3034 3035 3036
		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
3037 3038
		/* async page fault always cause vmexit */
		else if ((exit_code == SVM_EXIT_EXCP_BASE + PF_VECTOR) &&
3039
			 svm->vcpu.arch.exception.nested_apf != 0)
G
Gleb Natapov 已提交
3040
			vmexit = NESTED_EXIT_DONE;
3041 3042
		break;
	}
3043 3044 3045 3046
	case SVM_EXIT_ERR: {
		vmexit = NESTED_EXIT_DONE;
		break;
	}
3047 3048
	default: {
		u64 exit_bits = 1ULL << (exit_code - SVM_EXIT_INTR);
J
Joerg Roedel 已提交
3049
		if (svm->nested.intercept & exit_bits)
3050
			vmexit = NESTED_EXIT_DONE;
3051 3052 3053
	}
	}

3054 3055 3056 3057 3058 3059 3060 3061 3062 3063
	return vmexit;
}

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

	vmexit = nested_svm_intercept(svm);

	if (vmexit == NESTED_EXIT_DONE)
3064 3065 3066
		nested_svm_vmexit(svm);

	return vmexit;
3067 3068
}

3069 3070 3071 3072 3073
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;

3074
	dst->intercept_cr         = from->intercept_cr;
3075
	dst->intercept_dr         = from->intercept_dr;
3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095
	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;
3096
	dst->virt_ext              = from->virt_ext;
3097 3098
}

3099
static int nested_svm_vmexit(struct vcpu_svm *svm)
3100
{
3101
	struct vmcb *nested_vmcb;
3102
	struct vmcb *hsave = svm->nested.hsave;
J
Joerg Roedel 已提交
3103
	struct vmcb *vmcb = svm->vmcb;
3104
	struct page *page;
3105

3106 3107 3108 3109
	trace_kvm_nested_vmexit_inject(vmcb->control.exit_code,
				       vmcb->control.exit_info_1,
				       vmcb->control.exit_info_2,
				       vmcb->control.exit_int_info,
3110 3111
				       vmcb->control.exit_int_info_err,
				       KVM_ISA_SVM);
3112

3113
	nested_vmcb = nested_svm_map(svm, svm->nested.vmcb, &page);
3114 3115 3116
	if (!nested_vmcb)
		return 1;

3117 3118
	/* Exit Guest-Mode */
	leave_guest_mode(&svm->vcpu);
3119 3120
	svm->nested.vmcb = 0;

3121
	/* Give the current vmcb to the guest */
J
Joerg Roedel 已提交
3122 3123 3124 3125 3126 3127 3128 3129
	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;
3130
	nested_vmcb->save.efer   = svm->vcpu.arch.efer;
3131
	nested_vmcb->save.cr0    = kvm_read_cr0(&svm->vcpu);
3132
	nested_vmcb->save.cr3    = kvm_read_cr3(&svm->vcpu);
J
Joerg Roedel 已提交
3133
	nested_vmcb->save.cr2    = vmcb->save.cr2;
3134
	nested_vmcb->save.cr4    = svm->vcpu.arch.cr4;
3135
	nested_vmcb->save.rflags = kvm_get_rflags(&svm->vcpu);
J
Joerg Roedel 已提交
3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151
	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;
3152 3153 3154

	if (svm->nrips_enabled)
		nested_vmcb->control.next_rip  = vmcb->control.next_rip;
3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170

	/*
	 * 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
3171 3172 3173
	nested_vmcb->control.tlb_ctl           = 0;
	nested_vmcb->control.event_inj         = 0;
	nested_vmcb->control.event_inj_err     = 0;
3174 3175 3176 3177 3178 3179

	/* 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 */
3180
	copy_vmcb_control_area(vmcb, hsave);
3181

3182 3183
	kvm_clear_exception_queue(&svm->vcpu);
	kvm_clear_interrupt_queue(&svm->vcpu);
3184

3185 3186
	svm->nested.nested_cr3 = 0;

3187 3188 3189 3190 3191 3192 3193
	/* 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;
3194
	kvm_set_rflags(&svm->vcpu, hsave->save.rflags);
3195 3196 3197 3198 3199 3200 3201
	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 {
3202
		(void)kvm_set_cr3(&svm->vcpu, hsave->save.cr3);
3203 3204 3205 3206 3207 3208 3209 3210
	}
	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;

3211 3212
	mark_all_dirty(svm->vmcb);

3213
	nested_svm_unmap(page);
3214

3215
	nested_svm_uninit_mmu_context(&svm->vcpu);
3216 3217 3218 3219 3220
	kvm_mmu_reset_context(&svm->vcpu);
	kvm_mmu_load(&svm->vcpu);

	return 0;
}
A
Alexander Graf 已提交
3221

3222
static bool nested_svm_vmrun_msrpm(struct vcpu_svm *svm)
A
Alexander Graf 已提交
3223
{
3224 3225
	/*
	 * This function merges the msr permission bitmaps of kvm and the
G
Guo Chao 已提交
3226
	 * nested vmcb. It is optimized in that it only merges the parts where
3227 3228
	 * the kvm msr permission bitmap may contain zero bits
	 */
A
Alexander Graf 已提交
3229
	int i;
3230

3231 3232
	if (!(svm->nested.intercept & (1ULL << INTERCEPT_MSR_PROT)))
		return true;
3233

3234 3235 3236
	for (i = 0; i < MSRPM_OFFSETS; i++) {
		u32 value, p;
		u64 offset;
3237

3238 3239
		if (msrpm_offsets[i] == 0xffffffff)
			break;
A
Alexander Graf 已提交
3240

3241 3242
		p      = msrpm_offsets[i];
		offset = svm->nested.vmcb_msrpm + (p * 4);
3243

3244
		if (kvm_vcpu_read_guest(&svm->vcpu, offset, &value, 4))
3245 3246 3247 3248
			return false;

		svm->nested.msrpm[p] = svm->msrpm[p] | value;
	}
A
Alexander Graf 已提交
3249

3250
	svm->vmcb->control.msrpm_base_pa = __sme_set(__pa(svm->nested.msrpm));
3251 3252

	return true;
A
Alexander Graf 已提交
3253 3254
}

3255 3256 3257 3258 3259
static bool nested_vmcb_checks(struct vmcb *vmcb)
{
	if ((vmcb->control.intercept & (1ULL << INTERCEPT_VMRUN)) == 0)
		return false;

3260 3261 3262
	if (vmcb->control.asid == 0)
		return false;

3263 3264
	if ((vmcb->control.nested_ctl & SVM_NESTED_CTL_NP_ENABLE) &&
	    !npt_enabled)
3265 3266
		return false;

3267 3268 3269
	return true;
}

L
Ladi Prosek 已提交
3270 3271
static void enter_svm_guest_mode(struct vcpu_svm *svm, u64 vmcb_gpa,
				 struct vmcb *nested_vmcb, struct page *page)
A
Alexander Graf 已提交
3272
{
3273
	if (kvm_get_rflags(&svm->vcpu) & X86_EFLAGS_IF)
A
Alexander Graf 已提交
3274 3275 3276 3277
		svm->vcpu.arch.hflags |= HF_HIF_MASK;
	else
		svm->vcpu.arch.hflags &= ~HF_HIF_MASK;

3278
	if (nested_vmcb->control.nested_ctl & SVM_NESTED_CTL_NP_ENABLE) {
3279 3280 3281 3282 3283
		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
3284 3285 3286 3287 3288 3289 3290
	/* 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;
3291
	kvm_set_rflags(&svm->vcpu, nested_vmcb->save.rflags);
A
Alexander Graf 已提交
3292 3293 3294 3295 3296 3297
	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;
3298
	} else
3299
		(void)kvm_set_cr3(&svm->vcpu, nested_vmcb->save.cr3);
3300 3301 3302 3303

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

J
Joerg Roedel 已提交
3304
	svm->vmcb->save.cr2 = svm->vcpu.arch.cr2 = nested_vmcb->save.cr2;
A
Alexander Graf 已提交
3305 3306 3307
	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
3308

A
Alexander Graf 已提交
3309 3310 3311 3312 3313 3314 3315 3316
	/* 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;

3317
	svm->nested.vmcb_msrpm = nested_vmcb->control.msrpm_base_pa & ~0x0fffULL;
3318
	svm->nested.vmcb_iopm  = nested_vmcb->control.iopm_base_pa  & ~0x0fffULL;
A
Alexander Graf 已提交
3319

J
Joerg Roedel 已提交
3320
	/* cache intercepts */
3321
	svm->nested.intercept_cr         = nested_vmcb->control.intercept_cr;
3322
	svm->nested.intercept_dr         = nested_vmcb->control.intercept_dr;
J
Joerg Roedel 已提交
3323 3324 3325
	svm->nested.intercept_exceptions = nested_vmcb->control.intercept_exceptions;
	svm->nested.intercept            = nested_vmcb->control.intercept;

3326
	svm_flush_tlb(&svm->vcpu, true);
A
Alexander Graf 已提交
3327 3328 3329 3330 3331 3332
	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;

3333 3334
	if (svm->vcpu.arch.hflags & HF_VINTR_MASK) {
		/* We only want the cr8 intercept bits of the guest */
3335 3336
		clr_cr_intercept(svm, INTERCEPT_CR8_READ);
		clr_cr_intercept(svm, INTERCEPT_CR8_WRITE);
3337 3338
	}

3339
	/* We don't want to see VMMCALLs from a nested guest */
3340
	clr_intercept(svm, INTERCEPT_VMMCALL);
3341

3342
	svm->vmcb->control.virt_ext = nested_vmcb->control.virt_ext;
A
Alexander Graf 已提交
3343 3344 3345 3346 3347 3348
	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;

3349
	nested_svm_unmap(page);
3350

3351 3352 3353
	/* Enter Guest-Mode */
	enter_guest_mode(&svm->vcpu);

3354 3355 3356 3357 3358 3359
	/*
	 * Merge guest and host intercepts - must be called  with vcpu in
	 * guest-mode to take affect here
	 */
	recalc_intercepts(svm);

3360
	svm->nested.vmcb = vmcb_gpa;
3361

3362
	enable_gif(svm);
A
Alexander Graf 已提交
3363

3364
	mark_all_dirty(svm->vmcb);
L
Ladi Prosek 已提交
3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431
}

static bool nested_svm_vmrun(struct vcpu_svm *svm)
{
	struct vmcb *nested_vmcb;
	struct vmcb *hsave = svm->nested.hsave;
	struct vmcb *vmcb = svm->vmcb;
	struct page *page;
	u64 vmcb_gpa;

	vmcb_gpa = svm->vmcb->save.rax;

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

	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;
	}

	trace_kvm_nested_vmrun(svm->vmcb->save.rip, vmcb_gpa,
			       nested_vmcb->save.rip,
			       nested_vmcb->control.int_ctl,
			       nested_vmcb->control.event_inj,
			       nested_vmcb->control.nested_ctl);

	trace_kvm_nested_intercepts(nested_vmcb->control.intercept_cr & 0xffff,
				    nested_vmcb->control.intercept_cr >> 16,
				    nested_vmcb->control.intercept_exceptions,
				    nested_vmcb->control.intercept);

	/* Clear internal status */
	kvm_clear_exception_queue(&svm->vcpu);
	kvm_clear_interrupt_queue(&svm->vcpu);

	/*
	 * Save the old vmcb, so we don't need to pick what we save, but can
	 * restore everything when a VMEXIT occurs
	 */
	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;
	hsave->save.efer   = svm->vcpu.arch.efer;
	hsave->save.cr0    = kvm_read_cr0(&svm->vcpu);
	hsave->save.cr4    = svm->vcpu.arch.cr4;
	hsave->save.rflags = kvm_get_rflags(&svm->vcpu);
	hsave->save.rip    = kvm_rip_read(&svm->vcpu);
	hsave->save.rsp    = vmcb->save.rsp;
	hsave->save.rax    = vmcb->save.rax;
	if (npt_enabled)
		hsave->save.cr3    = vmcb->save.cr3;
	else
		hsave->save.cr3    = kvm_read_cr3(&svm->vcpu);

	copy_vmcb_control_area(hsave, vmcb);

	enter_svm_guest_mode(svm, vmcb_gpa, nested_vmcb, page);
3432

3433
	return true;
A
Alexander Graf 已提交
3434 3435
}

3436
static void nested_svm_vmloadsave(struct vmcb *from_vmcb, struct vmcb *to_vmcb)
3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451
{
	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
3452
static int vmload_interception(struct vcpu_svm *svm)
3453
{
3454
	struct vmcb *nested_vmcb;
3455
	struct page *page;
3456
	int ret;
3457

3458 3459 3460
	if (nested_svm_check_permissions(svm))
		return 1;

3461
	nested_vmcb = nested_svm_map(svm, svm->vmcb->save.rax, &page);
3462 3463 3464
	if (!nested_vmcb)
		return 1;

3465
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3466
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3467

3468
	nested_svm_vmloadsave(nested_vmcb, svm->vmcb);
3469
	nested_svm_unmap(page);
3470

3471
	return ret;
3472 3473
}

A
Avi Kivity 已提交
3474
static int vmsave_interception(struct vcpu_svm *svm)
3475
{
3476
	struct vmcb *nested_vmcb;
3477
	struct page *page;
3478
	int ret;
3479

3480 3481 3482
	if (nested_svm_check_permissions(svm))
		return 1;

3483
	nested_vmcb = nested_svm_map(svm, svm->vmcb->save.rax, &page);
3484 3485 3486
	if (!nested_vmcb)
		return 1;

3487
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3488
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3489

3490
	nested_svm_vmloadsave(svm->vmcb, nested_vmcb);
3491
	nested_svm_unmap(page);
3492

3493
	return ret;
3494 3495
}

A
Avi Kivity 已提交
3496
static int vmrun_interception(struct vcpu_svm *svm)
A
Alexander Graf 已提交
3497 3498 3499 3500
{
	if (nested_svm_check_permissions(svm))
		return 1;

3501 3502
	/* Save rip after vmrun instruction */
	kvm_rip_write(&svm->vcpu, kvm_rip_read(&svm->vcpu) + 3);
A
Alexander Graf 已提交
3503

3504
	if (!nested_svm_vmrun(svm))
A
Alexander Graf 已提交
3505 3506
		return 1;

3507
	if (!nested_svm_vmrun_msrpm(svm))
3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519
		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
3520 3521 3522 3523

	return 1;
}

A
Avi Kivity 已提交
3524
static int stgi_interception(struct vcpu_svm *svm)
3525
{
3526 3527
	int ret;

3528 3529 3530
	if (nested_svm_check_permissions(svm))
		return 1;

3531 3532
	/*
	 * If VGIF is enabled, the STGI intercept is only added to
3533
	 * detect the opening of the SMI/NMI window; remove it now.
3534 3535 3536 3537
	 */
	if (vgif_enabled(svm))
		clr_intercept(svm, INTERCEPT_STGI);

3538
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3539
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3540
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
3541

3542
	enable_gif(svm);
3543

3544
	return ret;
3545 3546
}

A
Avi Kivity 已提交
3547
static int clgi_interception(struct vcpu_svm *svm)
3548
{
3549 3550
	int ret;

3551 3552 3553 3554
	if (nested_svm_check_permissions(svm))
		return 1;

	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3555
	ret = kvm_skip_emulated_instruction(&svm->vcpu);
3556

3557
	disable_gif(svm);
3558 3559

	/* After a CLGI no interrupts should come */
3560 3561 3562 3563 3564
	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);
	}
3565

3566
	return ret;
3567 3568
}

A
Avi Kivity 已提交
3569
static int invlpga_interception(struct vcpu_svm *svm)
A
Alexander Graf 已提交
3570 3571 3572
{
	struct kvm_vcpu *vcpu = &svm->vcpu;

3573 3574
	trace_kvm_invlpga(svm->vmcb->save.rip, kvm_register_read(&svm->vcpu, VCPU_REGS_RCX),
			  kvm_register_read(&svm->vcpu, VCPU_REGS_RAX));
3575

A
Alexander Graf 已提交
3576
	/* Let's treat INVLPGA the same as INVLPG (can be optimized!) */
3577
	kvm_mmu_invlpg(vcpu, kvm_register_read(&svm->vcpu, VCPU_REGS_RAX));
A
Alexander Graf 已提交
3578 3579

	svm->next_rip = kvm_rip_read(&svm->vcpu) + 3;
3580
	return kvm_skip_emulated_instruction(&svm->vcpu);
A
Alexander Graf 已提交
3581 3582
}

3583 3584
static int skinit_interception(struct vcpu_svm *svm)
{
3585
	trace_kvm_skinit(svm->vmcb->save.rip, kvm_register_read(&svm->vcpu, VCPU_REGS_RAX));
3586 3587 3588 3589 3590

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

D
David Kaplan 已提交
3591 3592
static int wbinvd_interception(struct vcpu_svm *svm)
{
3593
	return kvm_emulate_wbinvd(&svm->vcpu);
D
David Kaplan 已提交
3594 3595
}

J
Joerg Roedel 已提交
3596 3597 3598 3599 3600 3601 3602
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;
3603
		return kvm_skip_emulated_instruction(&svm->vcpu);
J
Joerg Roedel 已提交
3604 3605 3606 3607 3608
	}

	return 1;
}

A
Avi Kivity 已提交
3609
static int task_switch_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3610
{
3611
	u16 tss_selector;
3612 3613 3614
	int reason;
	int int_type = svm->vmcb->control.exit_int_info &
		SVM_EXITINTINFO_TYPE_MASK;
3615
	int int_vec = svm->vmcb->control.exit_int_info & SVM_EVTINJ_VEC_MASK;
3616 3617 3618 3619
	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;
3620 3621
	bool has_error_code = false;
	u32 error_code = 0;
3622 3623

	tss_selector = (u16)svm->vmcb->control.exit_info_1;
3624

3625 3626
	if (svm->vmcb->control.exit_info_2 &
	    (1ULL << SVM_EXITINFOSHIFT_TS_REASON_IRET))
3627 3628 3629 3630
		reason = TASK_SWITCH_IRET;
	else if (svm->vmcb->control.exit_info_2 &
		 (1ULL << SVM_EXITINFOSHIFT_TS_REASON_JMP))
		reason = TASK_SWITCH_JMP;
3631
	else if (idt_v)
3632 3633 3634 3635
		reason = TASK_SWITCH_GATE;
	else
		reason = TASK_SWITCH_CALL;

3636 3637 3638 3639 3640 3641
	if (reason == TASK_SWITCH_GATE) {
		switch (type) {
		case SVM_EXITINTINFO_TYPE_NMI:
			svm->vcpu.arch.nmi_injected = false;
			break;
		case SVM_EXITINTINFO_TYPE_EXEPT:
3642 3643 3644 3645 3646 3647
			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;
			}
3648 3649 3650 3651 3652 3653 3654 3655 3656
			kvm_clear_exception_queue(&svm->vcpu);
			break;
		case SVM_EXITINTINFO_TYPE_INTR:
			kvm_clear_interrupt_queue(&svm->vcpu);
			break;
		default:
			break;
		}
	}
3657

3658 3659 3660
	if (reason != TASK_SWITCH_GATE ||
	    int_type == SVM_EXITINTINFO_TYPE_SOFT ||
	    (int_type == SVM_EXITINTINFO_TYPE_EXEPT &&
3661 3662
	     (int_vec == OF_VECTOR || int_vec == BP_VECTOR)))
		skip_emulated_instruction(&svm->vcpu);
3663

3664 3665 3666 3667
	if (int_type != SVM_EXITINTINFO_TYPE_SOFT)
		int_vec = -1;

	if (kvm_task_switch(&svm->vcpu, tss_selector, int_vec, reason,
3668 3669 3670 3671 3672 3673 3674
				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
3675 3676
}

A
Avi Kivity 已提交
3677
static int cpuid_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3678
{
3679
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 2;
3680
	return kvm_emulate_cpuid(&svm->vcpu);
A
Avi Kivity 已提交
3681 3682
}

A
Avi Kivity 已提交
3683
static int iret_interception(struct vcpu_svm *svm)
3684 3685
{
	++svm->vcpu.stat.nmi_window_exits;
3686
	clr_intercept(svm, INTERCEPT_IRET);
3687
	svm->vcpu.arch.hflags |= HF_IRET_MASK;
3688
	svm->nmi_iret_rip = kvm_rip_read(&svm->vcpu);
3689
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
3690 3691 3692
	return 1;
}

A
Avi Kivity 已提交
3693
static int invlpg_interception(struct vcpu_svm *svm)
M
Marcelo Tosatti 已提交
3694
{
3695 3696 3697 3698
	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);
3699
	return kvm_skip_emulated_instruction(&svm->vcpu);
M
Marcelo Tosatti 已提交
3700 3701
}

A
Avi Kivity 已提交
3702
static int emulate_on_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3703
{
3704
	return emulate_instruction(&svm->vcpu, 0) == EMULATE_DONE;
A
Avi Kivity 已提交
3705 3706
}

B
Brijesh Singh 已提交
3707 3708 3709 3710 3711 3712
static int rsm_interception(struct vcpu_svm *svm)
{
	return x86_emulate_instruction(&svm->vcpu, 0, 0,
				       rsm_ins_bytes, 2) == EMULATE_DONE;
}

A
Avi Kivity 已提交
3713 3714 3715 3716 3717 3718 3719 3720
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);
3721
	return kvm_complete_insn_gp(&svm->vcpu, err);
A
Avi Kivity 已提交
3722 3723
}

3724 3725
static bool check_selective_cr0_intercepted(struct vcpu_svm *svm,
					    unsigned long val)
3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747
{
	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;
}

3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762
#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;
3763 3764 3765 3766
	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;
3767 3768 3769 3770 3771 3772 3773

	err = 0;
	if (cr >= 16) { /* mov to cr */
		cr -= 16;
		val = kvm_register_read(&svm->vcpu, reg);
		switch (cr) {
		case 0:
3774 3775
			if (!check_selective_cr0_intercepted(svm, val))
				err = kvm_set_cr0(&svm->vcpu, val);
3776 3777 3778
			else
				return 1;

3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802
			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:
3803
			val = kvm_read_cr3(&svm->vcpu);
3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817
			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);
	}
3818
	return kvm_complete_insn_gp(&svm->vcpu, err);
3819 3820
}

3821 3822 3823 3824 3825
static int dr_interception(struct vcpu_svm *svm)
{
	int reg, dr;
	unsigned long val;

3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836
	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;
	}

3837 3838 3839 3840 3841 3842 3843
	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 */
3844 3845
		if (!kvm_require_dr(&svm->vcpu, dr - 16))
			return 1;
3846 3847 3848
		val = kvm_register_read(&svm->vcpu, reg);
		kvm_set_dr(&svm->vcpu, dr - 16, val);
	} else {
3849 3850 3851 3852
		if (!kvm_require_dr(&svm->vcpu, dr))
			return 1;
		kvm_get_dr(&svm->vcpu, dr, &val);
		kvm_register_write(&svm->vcpu, reg, val);
3853 3854
	}

3855
	return kvm_skip_emulated_instruction(&svm->vcpu);
3856 3857
}

A
Avi Kivity 已提交
3858
static int cr8_write_interception(struct vcpu_svm *svm)
3859
{
A
Avi Kivity 已提交
3860
	struct kvm_run *kvm_run = svm->vcpu.run;
A
Andre Przywara 已提交
3861
	int r;
A
Avi Kivity 已提交
3862

3863 3864
	u8 cr8_prev = kvm_get_cr8(&svm->vcpu);
	/* instruction emulation calls kvm_set_cr8() */
3865
	r = cr_interception(svm);
3866
	if (lapic_in_kernel(&svm->vcpu))
3867
		return r;
3868
	if (cr8_prev <= kvm_get_cr8(&svm->vcpu))
3869
		return r;
3870 3871 3872 3873
	kvm_run->exit_reason = KVM_EXIT_SET_TPR;
	return 0;
}

3874 3875
static int svm_get_msr_feature(struct kvm_msr_entry *msr)
{
3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887
	msr->data = 0;

	switch (msr->index) {
	case MSR_F10H_DECFG:
		if (boot_cpu_has(X86_FEATURE_LFENCE_RDTSC))
			msr->data |= MSR_F10H_DECFG_LFENCE_SERIALIZE;
		break;
	default:
		return 1;
	}

	return 0;
3888 3889
}

3890
static int svm_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
A
Avi Kivity 已提交
3891
{
3892 3893
	struct vcpu_svm *svm = to_svm(vcpu);

3894
	switch (msr_info->index) {
3895
	case MSR_IA32_TSC: {
3896
		msr_info->data = svm->vmcb->control.tsc_offset +
3897
			kvm_scale_tsc(vcpu, rdtsc());
3898

A
Avi Kivity 已提交
3899 3900
		break;
	}
B
Brian Gerst 已提交
3901
	case MSR_STAR:
3902
		msr_info->data = svm->vmcb->save.star;
A
Avi Kivity 已提交
3903
		break;
3904
#ifdef CONFIG_X86_64
A
Avi Kivity 已提交
3905
	case MSR_LSTAR:
3906
		msr_info->data = svm->vmcb->save.lstar;
A
Avi Kivity 已提交
3907 3908
		break;
	case MSR_CSTAR:
3909
		msr_info->data = svm->vmcb->save.cstar;
A
Avi Kivity 已提交
3910 3911
		break;
	case MSR_KERNEL_GS_BASE:
3912
		msr_info->data = svm->vmcb->save.kernel_gs_base;
A
Avi Kivity 已提交
3913 3914
		break;
	case MSR_SYSCALL_MASK:
3915
		msr_info->data = svm->vmcb->save.sfmask;
A
Avi Kivity 已提交
3916 3917 3918
		break;
#endif
	case MSR_IA32_SYSENTER_CS:
3919
		msr_info->data = svm->vmcb->save.sysenter_cs;
A
Avi Kivity 已提交
3920 3921
		break;
	case MSR_IA32_SYSENTER_EIP:
3922
		msr_info->data = svm->sysenter_eip;
A
Avi Kivity 已提交
3923 3924
		break;
	case MSR_IA32_SYSENTER_ESP:
3925
		msr_info->data = svm->sysenter_esp;
A
Avi Kivity 已提交
3926
		break;
P
Paolo Bonzini 已提交
3927 3928 3929 3930 3931
	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
3932 3933 3934 3935 3936
	/*
	 * 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.
	 */
3937
	case MSR_IA32_DEBUGCTLMSR:
3938
		msr_info->data = svm->vmcb->save.dbgctl;
3939 3940
		break;
	case MSR_IA32_LASTBRANCHFROMIP:
3941
		msr_info->data = svm->vmcb->save.br_from;
3942 3943
		break;
	case MSR_IA32_LASTBRANCHTOIP:
3944
		msr_info->data = svm->vmcb->save.br_to;
3945 3946
		break;
	case MSR_IA32_LASTINTFROMIP:
3947
		msr_info->data = svm->vmcb->save.last_excp_from;
3948 3949
		break;
	case MSR_IA32_LASTINTTOIP:
3950
		msr_info->data = svm->vmcb->save.last_excp_to;
3951
		break;
A
Alexander Graf 已提交
3952
	case MSR_VM_HSAVE_PA:
3953
		msr_info->data = svm->nested.hsave_msr;
A
Alexander Graf 已提交
3954
		break;
3955
	case MSR_VM_CR:
3956
		msr_info->data = svm->nested.vm_cr_msr;
3957
		break;
3958 3959 3960 3961 3962 3963 3964
	case MSR_IA32_SPEC_CTRL:
		if (!msr_info->host_initiated &&
		    !guest_cpuid_has(vcpu, X86_FEATURE_IBRS))
			return 1;

		msr_info->data = svm->spec_ctrl;
		break;
3965
	case MSR_IA32_UCODE_REV:
3966
		msr_info->data = 0x01000065;
3967
		break;
3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984
	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;
3985 3986 3987
	case MSR_F10H_DECFG:
		msr_info->data = svm->msr_decfg;
		break;
A
Avi Kivity 已提交
3988
	default:
3989
		return kvm_get_msr_common(vcpu, msr_info);
A
Avi Kivity 已提交
3990 3991 3992 3993
	}
	return 0;
}

A
Avi Kivity 已提交
3994
static int rdmsr_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
3995
{
3996
	u32 ecx = kvm_register_read(&svm->vcpu, VCPU_REGS_RCX);
3997
	struct msr_data msr_info;
A
Avi Kivity 已提交
3998

3999 4000 4001
	msr_info.index = ecx;
	msr_info.host_initiated = false;
	if (svm_get_msr(&svm->vcpu, &msr_info)) {
4002
		trace_kvm_msr_read_ex(ecx);
4003
		kvm_inject_gp(&svm->vcpu, 0);
4004
		return 1;
4005
	} else {
4006
		trace_kvm_msr_read(ecx, msr_info.data);
4007

4008 4009 4010 4011
		kvm_register_write(&svm->vcpu, VCPU_REGS_RAX,
				   msr_info.data & 0xffffffff);
		kvm_register_write(&svm->vcpu, VCPU_REGS_RDX,
				   msr_info.data >> 32);
4012
		svm->next_rip = kvm_rip_read(&svm->vcpu) + 2;
4013
		return kvm_skip_emulated_instruction(&svm->vcpu);
A
Avi Kivity 已提交
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
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;
}

4042
static int svm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr)
A
Avi Kivity 已提交
4043
{
4044 4045
	struct vcpu_svm *svm = to_svm(vcpu);

4046 4047
	u32 ecx = msr->index;
	u64 data = msr->data;
A
Avi Kivity 已提交
4048
	switch (ecx) {
P
Paolo Bonzini 已提交
4049 4050 4051 4052 4053 4054 4055
	case MSR_IA32_CR_PAT:
		if (!kvm_mtrr_valid(vcpu, MSR_IA32_CR_PAT, data))
			return 1;
		vcpu->arch.pat = data;
		svm->vmcb->save.g_pat = data;
		mark_dirty(svm->vmcb, VMCB_NPT);
		break;
4056
	case MSR_IA32_TSC:
4057
		kvm_write_tsc(vcpu, msr);
A
Avi Kivity 已提交
4058
		break;
4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085
	case MSR_IA32_SPEC_CTRL:
		if (!msr->host_initiated &&
		    !guest_cpuid_has(vcpu, X86_FEATURE_IBRS))
			return 1;

		/* The STIBP bit doesn't fault even if it's not advertised */
		if (data & ~(SPEC_CTRL_IBRS | SPEC_CTRL_STIBP))
			return 1;

		svm->spec_ctrl = data;

		if (!data)
			break;

		/*
		 * For non-nested:
		 * When it's written (to non-zero) for the first time, pass
		 * it through.
		 *
		 * For nested:
		 * The handling of the MSR bitmap for L2 guests is done in
		 * nested_svm_vmrun_msrpm.
		 * We update the L1 MSR bit as well since it will end up
		 * touching the MSR anyway now.
		 */
		set_msr_interception(svm->msrpm, MSR_IA32_SPEC_CTRL, 1, 1);
		break;
A
Ashok Raj 已提交
4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101
	case MSR_IA32_PRED_CMD:
		if (!msr->host_initiated &&
		    !guest_cpuid_has(vcpu, X86_FEATURE_IBPB))
			return 1;

		if (data & ~PRED_CMD_IBPB)
			return 1;

		if (!data)
			break;

		wrmsrl(MSR_IA32_PRED_CMD, PRED_CMD_IBPB);
		if (is_guest_mode(vcpu))
			break;
		set_msr_interception(svm->msrpm, MSR_IA32_PRED_CMD, 0, 1);
		break;
B
Brian Gerst 已提交
4102
	case MSR_STAR:
4103
		svm->vmcb->save.star = data;
A
Avi Kivity 已提交
4104
		break;
4105
#ifdef CONFIG_X86_64
A
Avi Kivity 已提交
4106
	case MSR_LSTAR:
4107
		svm->vmcb->save.lstar = data;
A
Avi Kivity 已提交
4108 4109
		break;
	case MSR_CSTAR:
4110
		svm->vmcb->save.cstar = data;
A
Avi Kivity 已提交
4111 4112
		break;
	case MSR_KERNEL_GS_BASE:
4113
		svm->vmcb->save.kernel_gs_base = data;
A
Avi Kivity 已提交
4114 4115
		break;
	case MSR_SYSCALL_MASK:
4116
		svm->vmcb->save.sfmask = data;
A
Avi Kivity 已提交
4117 4118 4119
		break;
#endif
	case MSR_IA32_SYSENTER_CS:
4120
		svm->vmcb->save.sysenter_cs = data;
A
Avi Kivity 已提交
4121 4122
		break;
	case MSR_IA32_SYSENTER_EIP:
4123
		svm->sysenter_eip = data;
4124
		svm->vmcb->save.sysenter_eip = data;
A
Avi Kivity 已提交
4125 4126
		break;
	case MSR_IA32_SYSENTER_ESP:
4127
		svm->sysenter_esp = data;
4128
		svm->vmcb->save.sysenter_esp = data;
A
Avi Kivity 已提交
4129
		break;
P
Paolo Bonzini 已提交
4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141
	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;
4142
	case MSR_IA32_DEBUGCTLMSR:
4143
		if (!boot_cpu_has(X86_FEATURE_LBRV)) {
4144 4145
			vcpu_unimpl(vcpu, "%s: MSR_IA32_DEBUGCTL 0x%llx, nop\n",
				    __func__, data);
4146 4147 4148 4149 4150 4151
			break;
		}
		if (data & DEBUGCTL_RESERVED_BITS)
			return 1;

		svm->vmcb->save.dbgctl = data;
4152
		mark_dirty(svm->vmcb, VMCB_LBR);
4153 4154 4155 4156
		if (data & (1ULL<<0))
			svm_enable_lbrv(svm);
		else
			svm_disable_lbrv(svm);
4157
		break;
A
Alexander Graf 已提交
4158
	case MSR_VM_HSAVE_PA:
4159
		svm->nested.hsave_msr = data;
4160
		break;
4161
	case MSR_VM_CR:
4162
		return svm_set_vm_cr(vcpu, data);
4163
	case MSR_VM_IGNNE:
4164
		vcpu_unimpl(vcpu, "unimplemented wrmsr: 0x%x data 0x%llx\n", ecx, data);
4165
		break;
4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183
	case MSR_F10H_DECFG: {
		struct kvm_msr_entry msr_entry;

		msr_entry.index = msr->index;
		if (svm_get_msr_feature(&msr_entry))
			return 1;

		/* Check the supported bits */
		if (data & ~msr_entry.data)
			return 1;

		/* Don't allow the guest to change a bit, #GP */
		if (!msr->host_initiated && (data ^ msr_entry.data))
			return 1;

		svm->msr_decfg = data;
		break;
	}
4184 4185 4186 4187
	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
4188
	default:
4189
		return kvm_set_msr_common(vcpu, msr);
A
Avi Kivity 已提交
4190 4191 4192 4193
	}
	return 0;
}

A
Avi Kivity 已提交
4194
static int wrmsr_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
4195
{
4196
	struct msr_data msr;
4197 4198
	u32 ecx = kvm_register_read(&svm->vcpu, VCPU_REGS_RCX);
	u64 data = kvm_read_edx_eax(&svm->vcpu);
4199

4200 4201 4202
	msr.data = data;
	msr.index = ecx;
	msr.host_initiated = false;
4203

4204
	svm->next_rip = kvm_rip_read(&svm->vcpu) + 2;
4205
	if (kvm_set_msr(&svm->vcpu, &msr)) {
4206
		trace_kvm_msr_write_ex(ecx, data);
4207
		kvm_inject_gp(&svm->vcpu, 0);
4208
		return 1;
4209 4210
	} else {
		trace_kvm_msr_write(ecx, data);
4211
		return kvm_skip_emulated_instruction(&svm->vcpu);
4212
	}
A
Avi Kivity 已提交
4213 4214
}

A
Avi Kivity 已提交
4215
static int msr_interception(struct vcpu_svm *svm)
A
Avi Kivity 已提交
4216
{
R
Rusty Russell 已提交
4217
	if (svm->vmcb->control.exit_info_1)
A
Avi Kivity 已提交
4218
		return wrmsr_interception(svm);
A
Avi Kivity 已提交
4219
	else
A
Avi Kivity 已提交
4220
		return rdmsr_interception(svm);
A
Avi Kivity 已提交
4221 4222
}

A
Avi Kivity 已提交
4223
static int interrupt_window_interception(struct vcpu_svm *svm)
4224
{
4225
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
4226
	svm_clear_vintr(svm);
4227
	svm->vmcb->control.int_ctl &= ~V_IRQ_MASK;
4228
	mark_dirty(svm->vmcb, VMCB_INTR);
4229
	++svm->vcpu.stat.irq_window_exits;
4230 4231 4232
	return 1;
}

4233 4234
static int pause_interception(struct vcpu_svm *svm)
{
4235 4236 4237 4238
	struct kvm_vcpu *vcpu = &svm->vcpu;
	bool in_kernel = (svm_get_cpl(vcpu) == 0);

	kvm_vcpu_on_spin(vcpu, in_kernel);
4239 4240 4241
	return 1;
}

4242 4243
static int nop_interception(struct vcpu_svm *svm)
{
4244
	return kvm_skip_emulated_instruction(&(svm->vcpu));
4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258
}

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);
}

4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270
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
4271
	u32 index = svm->vmcb->control.exit_info_2 & 0xFF;
4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 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 4530
	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;
}

4531
static int (*const svm_exit_handlers[])(struct vcpu_svm *svm) = {
4532 4533 4534 4535
	[SVM_EXIT_READ_CR0]			= cr_interception,
	[SVM_EXIT_READ_CR3]			= cr_interception,
	[SVM_EXIT_READ_CR4]			= cr_interception,
	[SVM_EXIT_READ_CR8]			= cr_interception,
4536
	[SVM_EXIT_CR0_SEL_WRITE]		= cr_interception,
4537
	[SVM_EXIT_WRITE_CR0]			= cr_interception,
4538 4539
	[SVM_EXIT_WRITE_CR3]			= cr_interception,
	[SVM_EXIT_WRITE_CR4]			= cr_interception,
J
Joerg Roedel 已提交
4540
	[SVM_EXIT_WRITE_CR8]			= cr8_write_interception,
4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556
	[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
4557 4558
	[SVM_EXIT_EXCP_BASE + DB_VECTOR]	= db_interception,
	[SVM_EXIT_EXCP_BASE + BP_VECTOR]	= bp_interception,
4559
	[SVM_EXIT_EXCP_BASE + UD_VECTOR]	= ud_interception,
J
Joerg Roedel 已提交
4560 4561
	[SVM_EXIT_EXCP_BASE + PF_VECTOR]	= pf_interception,
	[SVM_EXIT_EXCP_BASE + MC_VECTOR]	= mc_interception,
4562
	[SVM_EXIT_EXCP_BASE + AC_VECTOR]	= ac_interception,
J
Joerg Roedel 已提交
4563
	[SVM_EXIT_INTR]				= intr_interception,
4564
	[SVM_EXIT_NMI]				= nmi_interception,
A
Avi Kivity 已提交
4565 4566
	[SVM_EXIT_SMI]				= nop_on_interception,
	[SVM_EXIT_INIT]				= nop_on_interception,
4567
	[SVM_EXIT_VINTR]			= interrupt_window_interception,
A
Avi Kivity 已提交
4568
	[SVM_EXIT_RDPMC]			= rdpmc_interception,
A
Avi Kivity 已提交
4569
	[SVM_EXIT_CPUID]			= cpuid_interception,
4570
	[SVM_EXIT_IRET]                         = iret_interception,
4571
	[SVM_EXIT_INVD]                         = emulate_on_interception,
4572
	[SVM_EXIT_PAUSE]			= pause_interception,
A
Avi Kivity 已提交
4573
	[SVM_EXIT_HLT]				= halt_interception,
M
Marcelo Tosatti 已提交
4574
	[SVM_EXIT_INVLPG]			= invlpg_interception,
A
Alexander Graf 已提交
4575
	[SVM_EXIT_INVLPGA]			= invlpga_interception,
J
Joerg Roedel 已提交
4576
	[SVM_EXIT_IOIO]				= io_interception,
A
Avi Kivity 已提交
4577 4578
	[SVM_EXIT_MSR]				= msr_interception,
	[SVM_EXIT_TASK_SWITCH]			= task_switch_interception,
4579
	[SVM_EXIT_SHUTDOWN]			= shutdown_interception,
A
Alexander Graf 已提交
4580
	[SVM_EXIT_VMRUN]			= vmrun_interception,
4581
	[SVM_EXIT_VMMCALL]			= vmmcall_interception,
4582 4583
	[SVM_EXIT_VMLOAD]			= vmload_interception,
	[SVM_EXIT_VMSAVE]			= vmsave_interception,
4584 4585
	[SVM_EXIT_STGI]				= stgi_interception,
	[SVM_EXIT_CLGI]				= clgi_interception,
4586
	[SVM_EXIT_SKINIT]			= skinit_interception,
D
David Kaplan 已提交
4587
	[SVM_EXIT_WBINVD]                       = wbinvd_interception,
4588 4589
	[SVM_EXIT_MONITOR]			= monitor_interception,
	[SVM_EXIT_MWAIT]			= mwait_interception,
J
Joerg Roedel 已提交
4590
	[SVM_EXIT_XSETBV]			= xsetbv_interception,
4591
	[SVM_EXIT_NPF]				= npf_interception,
B
Brijesh Singh 已提交
4592
	[SVM_EXIT_RSM]                          = rsm_interception,
4593 4594
	[SVM_EXIT_AVIC_INCOMPLETE_IPI]		= avic_incomplete_ipi_interception,
	[SVM_EXIT_AVIC_UNACCELERATED_ACCESS]	= avic_unaccelerated_access_interception,
A
Avi Kivity 已提交
4595 4596
};

4597
static void dump_vmcb(struct kvm_vcpu *vcpu)
4598 4599 4600 4601 4602 4603
{
	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");
4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625
	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);
4626
	pr_err("%-20s%016llx\n", "avic_vapic_bar:", control->avic_vapic_bar);
4627 4628
	pr_err("%-20s%08x\n", "event_inj:", control->event_inj);
	pr_err("%-20s%08x\n", "event_inj_err:", control->event_inj_err);
4629
	pr_err("%-20s%lld\n", "virt_ext:", control->virt_ext);
4630
	pr_err("%-20s%016llx\n", "next_rip:", control->next_rip);
4631 4632 4633
	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);
4634
	pr_err("VMCB State Save Area:\n");
4635 4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674
	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);
4675 4676
	pr_err("cpl:            %d                efer:         %016llx\n",
		save->cpl, save->efer);
4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703
	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);
4704 4705
}

4706 4707 4708 4709 4710 4711 4712 4713
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
4714
static int handle_exit(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
4715
{
4716
	struct vcpu_svm *svm = to_svm(vcpu);
A
Avi Kivity 已提交
4717
	struct kvm_run *kvm_run = vcpu->run;
4718
	u32 exit_code = svm->vmcb->control.exit_code;
A
Avi Kivity 已提交
4719

4720 4721
	trace_kvm_exit(exit_code, vcpu, KVM_ISA_SVM);

4722
	if (!is_cr_intercept(svm, INTERCEPT_CR0_WRITE))
4723 4724 4725
		vcpu->arch.cr0 = svm->vmcb->save.cr0;
	if (npt_enabled)
		vcpu->arch.cr3 = svm->vmcb->save.cr3;
4726

4727 4728 4729 4730 4731 4732 4733
	if (unlikely(svm->nested.exit_required)) {
		nested_svm_vmexit(svm);
		svm->nested.exit_required = false;

		return 1;
	}

4734
	if (is_guest_mode(vcpu)) {
4735 4736
		int vmexit;

4737 4738 4739 4740
		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,
4741 4742
					svm->vmcb->control.exit_int_info_err,
					KVM_ISA_SVM);
4743

4744 4745 4746 4747 4748 4749
		vmexit = nested_svm_exit_special(svm);

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

		if (vmexit == NESTED_EXIT_DONE)
4750 4751 4752
			return 1;
	}

4753 4754
	svm_complete_interrupts(svm);

4755 4756 4757 4758
	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;
4759 4760
		pr_err("KVM: FAILED VMRUN WITH VMCB:\n");
		dump_vmcb(vcpu);
4761 4762 4763
		return 0;
	}

4764
	if (is_external_interrupt(svm->vmcb->control.exit_int_info) &&
4765
	    exit_code != SVM_EXIT_EXCP_BASE + PF_VECTOR &&
4766 4767
	    exit_code != SVM_EXIT_NPF && exit_code != SVM_EXIT_TASK_SWITCH &&
	    exit_code != SVM_EXIT_INTR && exit_code != SVM_EXIT_NMI)
4768
		printk(KERN_ERR "%s: unexpected exit_int_info 0x%x "
A
Avi Kivity 已提交
4769
		       "exit_code 0x%x\n",
4770
		       __func__, svm->vmcb->control.exit_int_info,
A
Avi Kivity 已提交
4771 4772
		       exit_code);

4773
	if (exit_code >= ARRAY_SIZE(svm_exit_handlers)
J
Joe Perches 已提交
4774
	    || !svm_exit_handlers[exit_code]) {
4775
		WARN_ONCE(1, "svm: unexpected exit reason 0x%x\n", exit_code);
4776 4777
		kvm_queue_exception(vcpu, UD_VECTOR);
		return 1;
A
Avi Kivity 已提交
4778 4779
	}

A
Avi Kivity 已提交
4780
	return svm_exit_handlers[exit_code](svm);
A
Avi Kivity 已提交
4781 4782 4783 4784 4785 4786
}

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

4787 4788
	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
4789 4790 4791
	load_TR_desc();
}

4792 4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815
static void pre_sev_run(struct vcpu_svm *svm, int cpu)
{
	struct svm_cpu_data *sd = per_cpu(svm_data, cpu);
	int asid = sev_get_asid(svm->vcpu.kvm);

	/* Assign the asid allocated with this SEV guest */
	svm->vmcb->control.asid = asid;

	/*
	 * Flush guest TLB:
	 *
	 * 1) when different VMCB for the same ASID is to be run on the same host CPU.
	 * 2) or this VMCB was executed on different host CPU in previous VMRUNs.
	 */
	if (sd->sev_vmcbs[asid] == svm->vmcb &&
	    svm->last_cpu == cpu)
		return;

	svm->last_cpu = cpu;
	sd->sev_vmcbs[asid] = svm->vmcb;
	svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ASID;
	mark_dirty(svm->vmcb, VMCB_ASID);
}

R
Rusty Russell 已提交
4816
static void pre_svm_run(struct vcpu_svm *svm)
A
Avi Kivity 已提交
4817 4818 4819
{
	int cpu = raw_smp_processor_id();

4820
	struct svm_cpu_data *sd = per_cpu(svm_data, cpu);
A
Avi Kivity 已提交
4821

4822 4823 4824
	if (sev_guest(svm->vcpu.kvm))
		return pre_sev_run(svm, cpu);

4825
	/* FIXME: handle wraparound of asid_generation */
4826 4827
	if (svm->asid_generation != sd->asid_generation)
		new_asid(svm, sd);
A
Avi Kivity 已提交
4828 4829
}

4830 4831 4832 4833 4834 4835
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;
4836
	set_intercept(svm, INTERCEPT_IRET);
4837 4838
	++vcpu->stat.nmi_injections;
}
A
Avi Kivity 已提交
4839

4840
static inline void svm_inject_irq(struct vcpu_svm *svm, int irq)
A
Avi Kivity 已提交
4841 4842 4843
{
	struct vmcb_control_area *control;

4844
	/* The following fields are ignored when AVIC is enabled */
R
Rusty Russell 已提交
4845
	control = &svm->vmcb->control;
4846
	control->int_vector = irq;
A
Avi Kivity 已提交
4847 4848 4849
	control->int_ctl &= ~V_INTR_PRIO_MASK;
	control->int_ctl |= V_IRQ_MASK |
		((/*control->int_vector >> 4*/ 0xf) << V_INTR_PRIO_SHIFT);
4850
	mark_dirty(svm->vmcb, VMCB_INTR);
A
Avi Kivity 已提交
4851 4852
}

4853
static void svm_set_irq(struct kvm_vcpu *vcpu)
E
Eddie Dong 已提交
4854 4855 4856
{
	struct vcpu_svm *svm = to_svm(vcpu);

4857
	BUG_ON(!(gif_set(svm)));
4858

4859 4860 4861
	trace_kvm_inj_virq(vcpu->arch.interrupt.nr);
	++vcpu->stat.irq_injections;

4862 4863
	svm->vmcb->control.event_inj = vcpu->arch.interrupt.nr |
		SVM_EVTINJ_VALID | SVM_EVTINJ_TYPE_INTR;
E
Eddie Dong 已提交
4864 4865
}

4866 4867 4868 4869 4870
static inline bool svm_nested_virtualize_tpr(struct kvm_vcpu *vcpu)
{
	return is_guest_mode(vcpu) && (vcpu->arch.hflags & HF_VINTR_MASK);
}

4871
static void update_cr8_intercept(struct kvm_vcpu *vcpu, int tpr, int irr)
4872 4873 4874
{
	struct vcpu_svm *svm = to_svm(vcpu);

4875 4876
	if (svm_nested_virtualize_tpr(vcpu) ||
	    kvm_vcpu_apicv_active(vcpu))
4877 4878
		return;

4879 4880
	clr_cr_intercept(svm, INTERCEPT_CR8_WRITE);

4881
	if (irr == -1)
4882 4883
		return;

4884
	if (tpr >= irr)
4885
		set_cr_intercept(svm, INTERCEPT_CR8_WRITE);
4886
}
4887

4888 4889 4890 4891 4892
static void svm_set_virtual_x2apic_mode(struct kvm_vcpu *vcpu, bool set)
{
	return;
}

4893
static bool svm_get_enable_apicv(struct kvm_vcpu *vcpu)
4894
{
4895
	return avic && irqchip_split(vcpu->kvm);
4896 4897 4898 4899
}

static void svm_hwapic_irr_update(struct kvm_vcpu *vcpu, int max_irr)
{
4900 4901
}

4902
static void svm_hwapic_isr_update(struct kvm_vcpu *vcpu, int max_isr)
4903
{
4904 4905
}

4906
/* Note: Currently only used by Hyper-V. */
4907
static void svm_refresh_apicv_exec_ctrl(struct kvm_vcpu *vcpu)
4908
{
4909 4910 4911
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb *vmcb = svm->vmcb;

4912
	if (!kvm_vcpu_apicv_active(&svm->vcpu))
4913 4914 4915 4916
		return;

	vmcb->control.int_ctl &= ~AVIC_ENABLE_MASK;
	mark_dirty(vmcb, VMCB_INTR);
4917 4918
}

4919
static void svm_load_eoi_exitmap(struct kvm_vcpu *vcpu, u64 *eoi_exit_bitmap)
4920 4921 4922 4923
{
	return;
}

4924 4925 4926 4927 4928 4929 4930
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,
4931
		       kvm_cpu_get_apicid(vcpu->cpu));
4932 4933 4934 4935
	else
		kvm_vcpu_wake_up(vcpu);
}

4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024
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);
5025
	vcpu_info->pi_desc_addr = __sme_set(page_to_phys((*svm)->avic_backing_page));
5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075
	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 */
5076 5077
			pi.base = __sme_set(page_to_phys(svm->avic_backing_page) &
					    AVIC_HPA_MASK);
5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139
			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;
}

5140 5141 5142 5143
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
5144 5145 5146 5147 5148 5149
	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;
5150 5151
}

J
Jan Kiszka 已提交
5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164
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;
5165
		set_intercept(svm, INTERCEPT_IRET);
J
Jan Kiszka 已提交
5166 5167
	} else {
		svm->vcpu.arch.hflags &= ~HF_NMI_MASK;
5168
		clr_intercept(svm, INTERCEPT_IRET);
J
Jan Kiszka 已提交
5169 5170 5171
	}
}

5172 5173 5174 5175
static int svm_interrupt_allowed(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb *vmcb = svm->vmcb;
5176 5177 5178 5179 5180 5181
	int ret;

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

5182
	ret = !!(kvm_get_rflags(vcpu) & X86_EFLAGS_IF);
5183

5184
	if (is_guest_mode(vcpu))
5185 5186 5187
		return ret && !(svm->vcpu.arch.hflags & HF_VINTR_MASK);

	return ret;
5188 5189
}

5190
static void enable_irq_window(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
5191
{
5192 5193
	struct vcpu_svm *svm = to_svm(vcpu);

5194 5195 5196
	if (kvm_vcpu_apicv_active(vcpu))
		return;

J
Joerg Roedel 已提交
5197 5198 5199 5200
	/*
	 * 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
5201 5202 5203
	 * 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
5204
	 */
5205
	if ((vgif_enabled(svm) || gif_set(svm)) && nested_svm_intr(svm)) {
5206 5207 5208
		svm_set_vintr(svm);
		svm_inject_irq(svm, 0x0);
	}
5209 5210
}

5211
static void enable_nmi_window(struct kvm_vcpu *vcpu)
5212
{
5213
	struct vcpu_svm *svm = to_svm(vcpu);
5214

5215 5216
	if ((svm->vcpu.arch.hflags & (HF_NMI_MASK | HF_IRET_MASK))
	    == HF_NMI_MASK)
5217
		return; /* IRET will cause a vm exit */
5218

5219 5220 5221
	if (!gif_set(svm)) {
		if (vgif_enabled(svm))
			set_intercept(svm, INTERCEPT_STGI);
5222
		return; /* STGI will cause a vm exit */
5223
	}
5224 5225 5226 5227

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

J
Joerg Roedel 已提交
5228 5229 5230 5231
	/*
	 * Something prevents NMI from been injected. Single step over possible
	 * problem (IRET or exception injection or interrupt shadow)
	 */
5232
	svm->nmi_singlestep_guest_rflags = svm_get_rflags(vcpu);
J
Jan Kiszka 已提交
5233
	svm->nmi_singlestep = true;
5234
	svm->vmcb->save.rflags |= (X86_EFLAGS_TF | X86_EFLAGS_RF);
5235 5236
}

5237 5238 5239 5240 5241
static int svm_set_tss_addr(struct kvm *kvm, unsigned int addr)
{
	return 0;
}

5242
static void svm_flush_tlb(struct kvm_vcpu *vcpu, bool invalidate_gpa)
5243
{
5244 5245 5246 5247 5248 5249
	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--;
5250 5251
}

5252 5253 5254 5255
static void svm_prepare_guest_switch(struct kvm_vcpu *vcpu)
{
}

5256 5257 5258 5259
static inline void sync_cr8_to_lapic(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

5260
	if (svm_nested_virtualize_tpr(vcpu))
5261 5262
		return;

5263
	if (!is_cr_intercept(svm, INTERCEPT_CR8_WRITE)) {
5264
		int cr8 = svm->vmcb->control.int_ctl & V_TPR_MASK;
5265
		kvm_set_cr8(vcpu, cr8);
5266 5267 5268
	}
}

5269 5270 5271 5272 5273
static inline void sync_lapic_to_cr8(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);
	u64 cr8;

5274 5275
	if (svm_nested_virtualize_tpr(vcpu) ||
	    kvm_vcpu_apicv_active(vcpu))
5276 5277
		return;

5278 5279 5280 5281 5282
	cr8 = kvm_get_cr8(vcpu);
	svm->vmcb->control.int_ctl &= ~V_TPR_MASK;
	svm->vmcb->control.int_ctl |= cr8 & V_TPR_MASK;
}

5283 5284 5285 5286 5287
static void svm_complete_interrupts(struct vcpu_svm *svm)
{
	u8 vector;
	int type;
	u32 exitintinfo = svm->vmcb->control.exit_int_info;
5288 5289 5290
	unsigned int3_injected = svm->int3_injected;

	svm->int3_injected = 0;
5291

5292 5293 5294 5295 5296 5297
	/*
	 * 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) {
5298
		svm->vcpu.arch.hflags &= ~(HF_NMI_MASK | HF_IRET_MASK);
5299 5300
		kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
	}
5301

5302 5303 5304 5305 5306 5307 5308
	svm->vcpu.arch.nmi_injected = false;
	kvm_clear_exception_queue(&svm->vcpu);
	kvm_clear_interrupt_queue(&svm->vcpu);

	if (!(exitintinfo & SVM_EXITINTINFO_VALID))
		return;

5309 5310
	kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);

5311 5312 5313 5314 5315 5316 5317 5318
	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:
5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329
		/*
		 * 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);
5330
			break;
5331
		}
5332 5333
		if (exitintinfo & SVM_EXITINTINFO_VALID_ERR) {
			u32 err = svm->vmcb->control.exit_int_info_err;
5334
			kvm_requeue_exception_e(&svm->vcpu, vector, err);
5335 5336

		} else
5337
			kvm_requeue_exception(&svm->vcpu, vector);
5338 5339
		break;
	case SVM_EXITINTINFO_TYPE_INTR:
5340
		kvm_queue_interrupt(&svm->vcpu, vector, false);
5341 5342 5343 5344 5345 5346
		break;
	default:
		break;
	}
}

A
Avi Kivity 已提交
5347 5348 5349 5350 5351 5352 5353 5354 5355 5356 5357
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
5358
static void svm_vcpu_run(struct kvm_vcpu *vcpu)
A
Avi Kivity 已提交
5359
{
5360
	struct vcpu_svm *svm = to_svm(vcpu);
5361

5362 5363 5364 5365
	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];

5366 5367 5368 5369 5370 5371 5372
	/*
	 * A vmexit emulation is required before the vcpu can be executed
	 * again.
	 */
	if (unlikely(svm->nested.exit_required))
		return;

5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388
	/*
	 * 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
5389
	pre_svm_run(svm);
A
Avi Kivity 已提交
5390

5391 5392
	sync_lapic_to_cr8(vcpu);

5393
	svm->vmcb->save.cr2 = vcpu->arch.cr2;
A
Avi Kivity 已提交
5394

5395 5396 5397
	clgi();

	local_irq_enable();
5398

5399 5400 5401 5402 5403 5404 5405 5406 5407
	/*
	 * If this vCPU has touched SPEC_CTRL, restore the guest's value if
	 * it's non-zero. Since vmentry is serialising on affected CPUs, there
	 * is no need to worry about the conditional branch over the wrmsr
	 * being speculatively taken.
	 */
	if (svm->spec_ctrl)
		wrmsrl(MSR_IA32_SPEC_CTRL, svm->spec_ctrl);

A
Avi Kivity 已提交
5408
	asm volatile (
A
Avi Kivity 已提交
5409 5410 5411 5412 5413 5414 5415
		"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"
5416
#ifdef CONFIG_X86_64
R
Rusty Russell 已提交
5417 5418 5419 5420 5421 5422 5423 5424
		"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
5425 5426 5427
#endif

		/* Enter guest mode */
A
Avi Kivity 已提交
5428 5429
		"push %%" _ASM_AX " \n\t"
		"mov %c[vmcb](%[svm]), %%" _ASM_AX " \n\t"
5430 5431 5432
		__ex(SVM_VMLOAD) "\n\t"
		__ex(SVM_VMRUN) "\n\t"
		__ex(SVM_VMSAVE) "\n\t"
A
Avi Kivity 已提交
5433
		"pop %%" _ASM_AX " \n\t"
A
Avi Kivity 已提交
5434 5435

		/* Save guest registers, load host registers */
A
Avi Kivity 已提交
5436 5437 5438 5439 5440 5441
		"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"
5442
#ifdef CONFIG_X86_64
R
Rusty Russell 已提交
5443 5444 5445 5446 5447 5448 5449 5450
		"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"
5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468 5469
#endif
		/*
		* Clear host registers marked as clobbered to prevent
		* speculative use.
		*/
		"xor %%" _ASM_BX ", %%" _ASM_BX " \n\t"
		"xor %%" _ASM_CX ", %%" _ASM_CX " \n\t"
		"xor %%" _ASM_DX ", %%" _ASM_DX " \n\t"
		"xor %%" _ASM_SI ", %%" _ASM_SI " \n\t"
		"xor %%" _ASM_DI ", %%" _ASM_DI " \n\t"
#ifdef CONFIG_X86_64
		"xor %%r8, %%r8 \n\t"
		"xor %%r9, %%r9 \n\t"
		"xor %%r10, %%r10 \n\t"
		"xor %%r11, %%r11 \n\t"
		"xor %%r12, %%r12 \n\t"
		"xor %%r13, %%r13 \n\t"
		"xor %%r14, %%r14 \n\t"
		"xor %%r15, %%r15 \n\t"
A
Avi Kivity 已提交
5470
#endif
A
Avi Kivity 已提交
5471
		"pop %%" _ASM_BP
A
Avi Kivity 已提交
5472
		:
R
Rusty Russell 已提交
5473
		: [svm]"a"(svm),
A
Avi Kivity 已提交
5474
		  [vmcb]"i"(offsetof(struct vcpu_svm, vmcb_pa)),
5475 5476 5477 5478 5479 5480
		  [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]))
5481
#ifdef CONFIG_X86_64
5482 5483 5484 5485 5486 5487 5488 5489
		  , [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
5490
#endif
5491 5492
		: "cc", "memory"
#ifdef CONFIG_X86_64
A
Avi Kivity 已提交
5493
		, "rbx", "rcx", "rdx", "rsi", "rdi"
5494
		, "r8", "r9", "r10", "r11" , "r12", "r13", "r14", "r15"
A
Avi Kivity 已提交
5495 5496
#else
		, "ebx", "ecx", "edx", "esi", "edi"
5497 5498
#endif
		);
A
Avi Kivity 已提交
5499

5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520
	/*
	 * We do not use IBRS in the kernel. If this vCPU has used the
	 * SPEC_CTRL MSR it may have left it on; save the value and
	 * turn it off. This is much more efficient than blindly adding
	 * it to the atomic save/restore list. Especially as the former
	 * (Saving guest MSRs on vmexit) doesn't even exist in KVM.
	 *
	 * For non-nested case:
	 * If the L01 MSR bitmap does not intercept the MSR, then we need to
	 * save it.
	 *
	 * For nested case:
	 * If the L02 MSR bitmap does not intercept the MSR, then we need to
	 * save it.
	 */
	if (!msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL))
		rdmsrl(MSR_IA32_SPEC_CTRL, svm->spec_ctrl);

	if (svm->spec_ctrl)
		wrmsrl(MSR_IA32_SPEC_CTRL, 0);

5521 5522 5523
	/* Eliminate branch target predictions from guest mode */
	vmexit_fill_RSB();

5524 5525 5526
#ifdef CONFIG_X86_64
	wrmsrl(MSR_GS_BASE, svm->host.gs_base);
#else
5527
	loadsegment(fs, svm->host.fs);
5528 5529 5530
#ifndef CONFIG_X86_32_LAZY_GS
	loadsegment(gs, svm->host.gs);
#endif
5531
#endif
A
Avi Kivity 已提交
5532 5533 5534

	reload_tss(vcpu);

5535 5536
	local_irq_disable();

5537 5538 5539 5540 5541
	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;

5542 5543 5544 5545 5546 5547 5548 5549 5550 5551
	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);

5552 5553
	sync_cr8_to_lapic(vcpu);

5554
	svm->next_rip = 0;
5555

5556 5557
	svm->vmcb->control.tlb_ctl = TLB_CONTROL_DO_NOTHING;

G
Gleb Natapov 已提交
5558 5559
	/* if exit due to PF check for async PF */
	if (svm->vmcb->control.exit_code == SVM_EXIT_EXCP_BASE + PF_VECTOR)
5560
		svm->vcpu.arch.apf.host_apf_reason = kvm_read_and_reset_pf_reason();
G
Gleb Natapov 已提交
5561

A
Avi Kivity 已提交
5562 5563 5564 5565
	if (npt_enabled) {
		vcpu->arch.regs_avail &= ~(1 << VCPU_EXREG_PDPTR);
		vcpu->arch.regs_dirty &= ~(1 << VCPU_EXREG_PDPTR);
	}
5566 5567 5568 5569 5570 5571 5572 5573

	/*
	 * 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);
5574 5575

	mark_all_clean(svm->vmcb);
A
Avi Kivity 已提交
5576
}
5577
STACK_FRAME_NON_STANDARD(svm_vcpu_run);
A
Avi Kivity 已提交
5578 5579 5580

static void svm_set_cr3(struct kvm_vcpu *vcpu, unsigned long root)
{
5581 5582
	struct vcpu_svm *svm = to_svm(vcpu);

5583
	svm->vmcb->save.cr3 = __sme_set(root);
5584
	mark_dirty(svm->vmcb, VMCB_CR);
5585
	svm_flush_tlb(vcpu, true);
A
Avi Kivity 已提交
5586 5587
}

5588 5589 5590 5591
static void set_tdp_cr3(struct kvm_vcpu *vcpu, unsigned long root)
{
	struct vcpu_svm *svm = to_svm(vcpu);

5592
	svm->vmcb->control.nested_cr3 = __sme_set(root);
5593
	mark_dirty(svm->vmcb, VMCB_NPT);
5594 5595

	/* Also sync guest cr3 here in case we live migrate */
5596
	svm->vmcb->save.cr3 = kvm_read_cr3(vcpu);
5597
	mark_dirty(svm->vmcb, VMCB_CR);
5598

5599
	svm_flush_tlb(vcpu, true);
5600 5601
}

A
Avi Kivity 已提交
5602 5603
static int is_disabled(void)
{
5604 5605 5606 5607 5608 5609
	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
5610 5611 5612
	return 0;
}

I
Ingo Molnar 已提交
5613 5614 5615 5616 5617 5618 5619 5620 5621 5622 5623
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
5624 5625 5626 5627 5628
static void svm_check_processor_compat(void *rtn)
{
	*(int *)rtn = 0;
}

5629 5630 5631 5632 5633
static bool svm_cpu_has_accelerated_tpr(void)
{
	return false;
}

5634 5635 5636 5637 5638
static bool svm_has_high_real_mode_segbase(void)
{
	return true;
}

5639 5640 5641 5642 5643
static u64 svm_get_mt_mask(struct kvm_vcpu *vcpu, gfn_t gfn, bool is_mmio)
{
	return 0;
}

5644 5645
static void svm_cpuid_update(struct kvm_vcpu *vcpu)
{
5646 5647 5648
	struct vcpu_svm *svm = to_svm(vcpu);

	/* Update nrips enabled cache */
5649
	svm->nrips_enabled = !!guest_cpuid_has(&svm->vcpu, X86_FEATURE_NRIPS);
5650 5651 5652 5653

	if (!kvm_vcpu_apicv_active(vcpu))
		return;

5654
	guest_cpuid_clear(vcpu, X86_FEATURE_X2APIC);
5655 5656
}

5657 5658
static void svm_set_supported_cpuid(u32 func, struct kvm_cpuid_entry2 *entry)
{
5659
	switch (func) {
5660 5661 5662 5663
	case 0x1:
		if (avic)
			entry->ecx &= ~bit(X86_FEATURE_X2APIC);
		break;
5664 5665 5666 5667
	case 0x80000001:
		if (nested)
			entry->ecx |= (1 << 2); /* Set SVM bit */
		break;
5668 5669 5670 5671 5672
	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 */
5673 5674 5675 5676
		entry->edx = 0; /* Per default do not support any
				   additional features */

		/* Support next_rip if host supports it */
5677
		if (boot_cpu_has(X86_FEATURE_NRIPS))
5678
			entry->edx |= SVM_FEATURE_NRIP;
5679

5680 5681 5682 5683
		/* Support NPT for the guest if enabled */
		if (npt_enabled)
			entry->edx |= SVM_FEATURE_NPT;

5684
		break;
5685 5686 5687 5688 5689 5690
	case 0x8000001F:
		/* Support memory encryption cpuid if host supports it */
		if (boot_cpu_has(X86_FEATURE_SEV))
			cpuid(0x8000001f, &entry->eax, &entry->ebx,
				&entry->ecx, &entry->edx);

5691
	}
5692 5693
}

5694
static int svm_get_lpage_level(void)
5695
{
5696
	return PT_PDPE_LEVEL;
5697 5698
}

5699 5700
static bool svm_rdtscp_supported(void)
{
P
Paolo Bonzini 已提交
5701
	return boot_cpu_has(X86_FEATURE_RDTSCP);
5702 5703
}

5704 5705 5706 5707 5708
static bool svm_invpcid_supported(void)
{
	return false;
}

5709 5710 5711 5712 5713
static bool svm_mpx_supported(void)
{
	return false;
}

5714 5715 5716 5717 5718
static bool svm_xsaves_supported(void)
{
	return false;
}

5719 5720 5721 5722 5723
static bool svm_umip_emulated(void)
{
	return false;
}

5724 5725 5726 5727 5728
static bool svm_has_wbinvd_exit(void)
{
	return true;
}

5729
#define PRE_EX(exit)  { .exit_code = (exit), \
5730
			.stage = X86_ICPT_PRE_EXCEPT, }
5731
#define POST_EX(exit) { .exit_code = (exit), \
5732
			.stage = X86_ICPT_POST_EXCEPT, }
5733
#define POST_MEM(exit) { .exit_code = (exit), \
5734
			.stage = X86_ICPT_POST_MEMACCESS, }
5735

5736
static const struct __x86_intercept {
5737 5738 5739 5740 5741 5742 5743 5744
	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),
5745 5746
	[x86_intercept_dr_read]		= POST_EX(SVM_EXIT_READ_DR0),
	[x86_intercept_dr_write]	= POST_EX(SVM_EXIT_WRITE_DR0),
5747 5748 5749 5750 5751 5752 5753 5754
	[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),
5755 5756 5757 5758 5759 5760 5761 5762
	[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),
5763 5764 5765
	[x86_intercept_rdtscp]		= POST_EX(SVM_EXIT_RDTSCP),
	[x86_intercept_monitor]		= POST_MEM(SVM_EXIT_MONITOR),
	[x86_intercept_mwait]		= POST_EX(SVM_EXIT_MWAIT),
5766 5767 5768 5769 5770 5771 5772 5773 5774
	[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),
5775 5776 5777 5778 5779 5780 5781
	[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),
5782 5783 5784 5785
	[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),
5786 5787
};

5788
#undef PRE_EX
5789
#undef POST_EX
5790
#undef POST_MEM
5791

5792 5793 5794 5795
static int svm_check_intercept(struct kvm_vcpu *vcpu,
			       struct x86_instruction_info *info,
			       enum x86_intercept_stage stage)
{
5796 5797 5798 5799 5800 5801 5802 5803 5804 5805
	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];

5806
	if (stage != icpt_info.stage)
5807 5808 5809 5810 5811 5812 5813 5814 5815 5816 5817 5818 5819 5820
		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;

5821 5822
		if (icpt_info.exit_code != SVM_EXIT_WRITE_CR0 ||
		    info->intercept == x86_intercept_clts)
5823 5824 5825 5826 5827 5828 5829 5830 5831 5832 5833 5834 5835 5836 5837 5838 5839 5840 5841 5842 5843 5844 5845
			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;
	}
5846 5847 5848 5849
	case SVM_EXIT_READ_DR0:
	case SVM_EXIT_WRITE_DR0:
		icpt_info.exit_code += info->modrm_reg;
		break;
5850 5851 5852 5853 5854 5855
	case SVM_EXIT_MSR:
		if (info->intercept == x86_intercept_wrmsr)
			vmcb->control.exit_info_1 = 1;
		else
			vmcb->control.exit_info_1 = 0;
		break;
5856 5857 5858 5859 5860 5861 5862
	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;
5863
		break;
5864 5865 5866 5867 5868 5869
	case SVM_EXIT_IOIO: {
		u64 exit_info;
		u32 bytes;

		if (info->intercept == x86_intercept_in ||
		    info->intercept == x86_intercept_ins) {
5870 5871
			exit_info = ((info->src_val & 0xffff) << 16) |
				SVM_IOIO_TYPE_MASK;
5872
			bytes = info->dst_bytes;
5873
		} else {
5874
			exit_info = (info->dst_val & 0xffff) << 16;
5875
			bytes = info->src_bytes;
5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891 5892 5893 5894 5895
		}

		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;
	}
5896 5897 5898 5899
	default:
		break;
	}

5900 5901 5902
	/* TODO: Advertise NRIPS to guest hypervisor unconditionally */
	if (static_cpu_has(X86_FEATURE_NRIPS))
		vmcb->control.next_rip  = info->next_rip;
5903 5904 5905 5906 5907 5908 5909 5910
	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;
5911 5912
}

5913 5914 5915
static void svm_handle_external_intr(struct kvm_vcpu *vcpu)
{
	local_irq_enable();
P
Paolo Bonzini 已提交
5916 5917 5918 5919 5920 5921
	/*
	 * We must have an instruction with interrupts enabled, so
	 * the timer interrupt isn't delayed by the interrupt shadow.
	 */
	asm("nop");
	local_irq_disable();
5922 5923
}

5924 5925 5926 5927
static void svm_sched_in(struct kvm_vcpu *vcpu, int cpu)
{
}

5928 5929 5930 5931 5932 5933 5934 5935 5936
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);
}

5937 5938 5939 5940 5941 5942
static void svm_setup_mce(struct kvm_vcpu *vcpu)
{
	/* [63:9] are reserved. */
	vcpu->arch.mcg_cap &= 0x1ff;
}

5943 5944
static int svm_smi_allowed(struct kvm_vcpu *vcpu)
{
5945 5946 5947 5948 5949 5950 5951 5952 5953 5954 5955 5956 5957 5958
	struct vcpu_svm *svm = to_svm(vcpu);

	/* Per APM Vol.2 15.22.2 "Response to SMI" */
	if (!gif_set(svm))
		return 0;

	if (is_guest_mode(&svm->vcpu) &&
	    svm->nested.intercept & (1ULL << INTERCEPT_SMI)) {
		/* TODO: Might need to set exit_info_1 and exit_info_2 here */
		svm->vmcb->control.exit_code = SVM_EXIT_SMI;
		svm->nested.exit_required = true;
		return 0;
	}

5959 5960 5961
	return 1;
}

5962 5963
static int svm_pre_enter_smm(struct kvm_vcpu *vcpu, char *smstate)
{
5964 5965 5966 5967 5968 5969 5970 5971 5972 5973 5974 5975 5976 5977 5978 5979 5980
	struct vcpu_svm *svm = to_svm(vcpu);
	int ret;

	if (is_guest_mode(vcpu)) {
		/* FED8h - SVM Guest */
		put_smstate(u64, smstate, 0x7ed8, 1);
		/* FEE0h - SVM Guest VMCB Physical Address */
		put_smstate(u64, smstate, 0x7ee0, svm->nested.vmcb);

		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];

		ret = nested_svm_vmexit(svm);
		if (ret)
			return ret;
	}
5981 5982 5983 5984 5985
	return 0;
}

static int svm_pre_leave_smm(struct kvm_vcpu *vcpu, u64 smbase)
{
5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000 6001 6002 6003 6004 6005 6006 6007 6008 6009
	struct vcpu_svm *svm = to_svm(vcpu);
	struct vmcb *nested_vmcb;
	struct page *page;
	struct {
		u64 guest;
		u64 vmcb;
	} svm_state_save;
	int ret;

	ret = kvm_vcpu_read_guest(vcpu, smbase + 0xfed8, &svm_state_save,
				  sizeof(svm_state_save));
	if (ret)
		return ret;

	if (svm_state_save.guest) {
		vcpu->arch.hflags &= ~HF_SMM_MASK;
		nested_vmcb = nested_svm_map(svm, svm_state_save.vmcb, &page);
		if (nested_vmcb)
			enter_svm_guest_mode(svm, svm_state_save.vmcb, nested_vmcb, page);
		else
			ret = 1;
		vcpu->arch.hflags |= HF_SMM_MASK;
	}
	return ret;
6010 6011
}

6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022 6023 6024
static int enable_smi_window(struct kvm_vcpu *vcpu)
{
	struct vcpu_svm *svm = to_svm(vcpu);

	if (!gif_set(svm)) {
		if (vgif_enabled(svm))
			set_intercept(svm, INTERCEPT_STGI);
		/* STGI will cause a vm exit */
		return 1;
	}
	return 0;
}

B
Brijesh Singh 已提交
6025 6026 6027 6028 6029 6030 6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043 6044 6045 6046 6047 6048 6049 6050 6051 6052 6053 6054 6055
static int sev_asid_new(void)
{
	int pos;

	/*
	 * SEV-enabled guest must use asid from min_sev_asid to max_sev_asid.
	 */
	pos = find_next_zero_bit(sev_asid_bitmap, max_sev_asid, min_sev_asid - 1);
	if (pos >= max_sev_asid)
		return -EBUSY;

	set_bit(pos, sev_asid_bitmap);
	return pos + 1;
}

static int sev_guest_init(struct kvm *kvm, struct kvm_sev_cmd *argp)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	int asid, ret;

	ret = -EBUSY;
	asid = sev_asid_new();
	if (asid < 0)
		return ret;

	ret = sev_platform_init(&argp->error);
	if (ret)
		goto e_free;

	sev->active = true;
	sev->asid = asid;
6056
	INIT_LIST_HEAD(&sev->regions_list);
B
Brijesh Singh 已提交
6057 6058 6059 6060 6061 6062 6063 6064

	return 0;

e_free:
	__sev_asid_free(asid);
	return ret;
}

6065 6066 6067 6068 6069 6070 6071 6072 6073 6074 6075 6076 6077 6078 6079 6080 6081 6082 6083 6084 6085 6086 6087 6088 6089
static int sev_bind_asid(struct kvm *kvm, unsigned int handle, int *error)
{
	struct sev_data_activate *data;
	int asid = sev_get_asid(kvm);
	int ret;

	wbinvd_on_all_cpus();

	ret = sev_guest_df_flush(error);
	if (ret)
		return ret;

	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		return -ENOMEM;

	/* activate ASID on the given handle */
	data->handle = handle;
	data->asid   = asid;
	ret = sev_guest_activate(data, error);
	kfree(data);

	return ret;
}

6090
static int __sev_issue_cmd(int fd, int id, void *data, int *error)
6091 6092 6093 6094 6095 6096 6097 6098 6099 6100 6101 6102 6103 6104
{
	struct fd f;
	int ret;

	f = fdget(fd);
	if (!f.file)
		return -EBADF;

	ret = sev_issue_cmd_external_user(f.file, id, data, error);

	fdput(f);
	return ret;
}

6105 6106 6107 6108 6109 6110 6111
static int sev_issue_cmd(struct kvm *kvm, int id, void *data, int *error)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;

	return __sev_issue_cmd(sev->fd, id, data, error);
}

6112 6113 6114 6115 6116 6117 6118 6119 6120 6121 6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138 6139 6140 6141 6142 6143 6144 6145 6146 6147 6148 6149 6150 6151 6152 6153 6154 6155 6156 6157 6158
static int sev_launch_start(struct kvm *kvm, struct kvm_sev_cmd *argp)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct sev_data_launch_start *start;
	struct kvm_sev_launch_start params;
	void *dh_blob, *session_blob;
	int *error = &argp->error;
	int ret;

	if (!sev_guest(kvm))
		return -ENOTTY;

	if (copy_from_user(&params, (void __user *)(uintptr_t)argp->data, sizeof(params)))
		return -EFAULT;

	start = kzalloc(sizeof(*start), GFP_KERNEL);
	if (!start)
		return -ENOMEM;

	dh_blob = NULL;
	if (params.dh_uaddr) {
		dh_blob = psp_copy_user_blob(params.dh_uaddr, params.dh_len);
		if (IS_ERR(dh_blob)) {
			ret = PTR_ERR(dh_blob);
			goto e_free;
		}

		start->dh_cert_address = __sme_set(__pa(dh_blob));
		start->dh_cert_len = params.dh_len;
	}

	session_blob = NULL;
	if (params.session_uaddr) {
		session_blob = psp_copy_user_blob(params.session_uaddr, params.session_len);
		if (IS_ERR(session_blob)) {
			ret = PTR_ERR(session_blob);
			goto e_free_dh;
		}

		start->session_address = __sme_set(__pa(session_blob));
		start->session_len = params.session_len;
	}

	start->handle = params.handle;
	start->policy = params.policy;

	/* create memory encryption context */
6159
	ret = __sev_issue_cmd(argp->sev_fd, SEV_CMD_LAUNCH_START, start, error);
6160 6161 6162 6163 6164 6165 6166 6167 6168 6169 6170 6171 6172 6173 6174 6175 6176 6177 6178 6179 6180 6181 6182 6183 6184 6185 6186 6187
	if (ret)
		goto e_free_session;

	/* Bind ASID to this guest */
	ret = sev_bind_asid(kvm, start->handle, error);
	if (ret)
		goto e_free_session;

	/* return handle to userspace */
	params.handle = start->handle;
	if (copy_to_user((void __user *)(uintptr_t)argp->data, &params, sizeof(params))) {
		sev_unbind_asid(kvm, start->handle);
		ret = -EFAULT;
		goto e_free_session;
	}

	sev->handle = start->handle;
	sev->fd = argp->sev_fd;

e_free_session:
	kfree(session_blob);
e_free_dh:
	kfree(dh_blob);
e_free:
	kfree(start);
	return ret;
}

6188 6189 6190 6191 6192 6193 6194 6195 6196 6197 6198 6199 6200 6201 6202 6203 6204 6205 6206 6207 6208 6209 6210 6211 6212 6213 6214 6215 6216 6217 6218 6219 6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231 6232 6233 6234 6235 6236 6237 6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248 6249 6250 6251 6252 6253 6254 6255 6256 6257 6258 6259 6260 6261 6262 6263 6264 6265 6266 6267 6268 6269 6270 6271 6272 6273 6274 6275 6276 6277 6278 6279 6280 6281 6282 6283 6284
static int get_num_contig_pages(int idx, struct page **inpages,
				unsigned long npages)
{
	unsigned long paddr, next_paddr;
	int i = idx + 1, pages = 1;

	/* find the number of contiguous pages starting from idx */
	paddr = __sme_page_pa(inpages[idx]);
	while (i < npages) {
		next_paddr = __sme_page_pa(inpages[i++]);
		if ((paddr + PAGE_SIZE) == next_paddr) {
			pages++;
			paddr = next_paddr;
			continue;
		}
		break;
	}

	return pages;
}

static int sev_launch_update_data(struct kvm *kvm, struct kvm_sev_cmd *argp)
{
	unsigned long vaddr, vaddr_end, next_vaddr, npages, size;
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct kvm_sev_launch_update_data params;
	struct sev_data_launch_update_data *data;
	struct page **inpages;
	int i, ret, pages;

	if (!sev_guest(kvm))
		return -ENOTTY;

	if (copy_from_user(&params, (void __user *)(uintptr_t)argp->data, sizeof(params)))
		return -EFAULT;

	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		return -ENOMEM;

	vaddr = params.uaddr;
	size = params.len;
	vaddr_end = vaddr + size;

	/* Lock the user memory. */
	inpages = sev_pin_memory(kvm, vaddr, size, &npages, 1);
	if (!inpages) {
		ret = -ENOMEM;
		goto e_free;
	}

	/*
	 * The LAUNCH_UPDATE command will perform in-place encryption of the
	 * memory content (i.e it will write the same memory region with C=1).
	 * It's possible that the cache may contain the data with C=0, i.e.,
	 * unencrypted so invalidate it first.
	 */
	sev_clflush_pages(inpages, npages);

	for (i = 0; vaddr < vaddr_end; vaddr = next_vaddr, i += pages) {
		int offset, len;

		/*
		 * If the user buffer is not page-aligned, calculate the offset
		 * within the page.
		 */
		offset = vaddr & (PAGE_SIZE - 1);

		/* Calculate the number of pages that can be encrypted in one go. */
		pages = get_num_contig_pages(i, inpages, npages);

		len = min_t(size_t, ((pages * PAGE_SIZE) - offset), size);

		data->handle = sev->handle;
		data->len = len;
		data->address = __sme_page_pa(inpages[i]) + offset;
		ret = sev_issue_cmd(kvm, SEV_CMD_LAUNCH_UPDATE_DATA, data, &argp->error);
		if (ret)
			goto e_unpin;

		size -= len;
		next_vaddr = vaddr + len;
	}

e_unpin:
	/* content of memory is updated, mark pages dirty */
	for (i = 0; i < npages; i++) {
		set_page_dirty_lock(inpages[i]);
		mark_page_accessed(inpages[i]);
	}
	/* unlock the user pages */
	sev_unpin_memory(kvm, inpages, npages);
e_free:
	kfree(data);
	return ret;
}

6285 6286
static int sev_launch_measure(struct kvm *kvm, struct kvm_sev_cmd *argp)
{
6287
	void __user *measure = (void __user *)(uintptr_t)argp->data;
6288 6289 6290
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct sev_data_launch_measure *data;
	struct kvm_sev_launch_measure params;
6291
	void __user *p = NULL;
6292 6293 6294 6295 6296 6297
	void *blob = NULL;
	int ret;

	if (!sev_guest(kvm))
		return -ENOTTY;

6298
	if (copy_from_user(&params, measure, sizeof(params)))
6299 6300 6301 6302 6303 6304 6305 6306 6307 6308
		return -EFAULT;

	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		return -ENOMEM;

	/* User wants to query the blob length */
	if (!params.len)
		goto cmd;

6309 6310
	p = (void __user *)(uintptr_t)params.uaddr;
	if (p) {
6311 6312 6313 6314 6315 6316 6317 6318 6319 6320 6321 6322 6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334 6335 6336 6337 6338
		if (params.len > SEV_FW_BLOB_MAX_SIZE) {
			ret = -EINVAL;
			goto e_free;
		}

		ret = -ENOMEM;
		blob = kmalloc(params.len, GFP_KERNEL);
		if (!blob)
			goto e_free;

		data->address = __psp_pa(blob);
		data->len = params.len;
	}

cmd:
	data->handle = sev->handle;
	ret = sev_issue_cmd(kvm, SEV_CMD_LAUNCH_MEASURE, data, &argp->error);

	/*
	 * If we query the session length, FW responded with expected data.
	 */
	if (!params.len)
		goto done;

	if (ret)
		goto e_free_blob;

	if (blob) {
6339
		if (copy_to_user(p, blob, params.len))
6340 6341 6342 6343 6344
			ret = -EFAULT;
	}

done:
	params.len = data->len;
6345
	if (copy_to_user(measure, &params, sizeof(params)))
6346 6347 6348 6349 6350 6351 6352 6353
		ret = -EFAULT;
e_free_blob:
	kfree(blob);
e_free:
	kfree(data);
	return ret;
}

6354 6355 6356 6357 6358 6359 6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373
static int sev_launch_finish(struct kvm *kvm, struct kvm_sev_cmd *argp)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct sev_data_launch_finish *data;
	int ret;

	if (!sev_guest(kvm))
		return -ENOTTY;

	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		return -ENOMEM;

	data->handle = sev->handle;
	ret = sev_issue_cmd(kvm, SEV_CMD_LAUNCH_FINISH, data, &argp->error);

	kfree(data);
	return ret;
}

6374 6375 6376 6377 6378 6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389 6390 6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401 6402 6403
static int sev_guest_status(struct kvm *kvm, struct kvm_sev_cmd *argp)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct kvm_sev_guest_status params;
	struct sev_data_guest_status *data;
	int ret;

	if (!sev_guest(kvm))
		return -ENOTTY;

	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		return -ENOMEM;

	data->handle = sev->handle;
	ret = sev_issue_cmd(kvm, SEV_CMD_GUEST_STATUS, data, &argp->error);
	if (ret)
		goto e_free;

	params.policy = data->policy;
	params.state = data->state;
	params.handle = data->handle;

	if (copy_to_user((void __user *)(uintptr_t)argp->data, &params, sizeof(params)))
		ret = -EFAULT;
e_free:
	kfree(data);
	return ret;
}

6404 6405 6406 6407 6408 6409 6410 6411 6412 6413 6414 6415 6416 6417 6418 6419 6420 6421 6422 6423 6424 6425 6426 6427 6428 6429 6430 6431 6432 6433 6434 6435 6436 6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460 6461 6462 6463 6464 6465 6466 6467 6468 6469 6470 6471 6472 6473 6474 6475 6476 6477 6478 6479 6480
static int __sev_issue_dbg_cmd(struct kvm *kvm, unsigned long src,
			       unsigned long dst, int size,
			       int *error, bool enc)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct sev_data_dbg *data;
	int ret;

	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		return -ENOMEM;

	data->handle = sev->handle;
	data->dst_addr = dst;
	data->src_addr = src;
	data->len = size;

	ret = sev_issue_cmd(kvm,
			    enc ? SEV_CMD_DBG_ENCRYPT : SEV_CMD_DBG_DECRYPT,
			    data, error);
	kfree(data);
	return ret;
}

static int __sev_dbg_decrypt(struct kvm *kvm, unsigned long src_paddr,
			     unsigned long dst_paddr, int sz, int *err)
{
	int offset;

	/*
	 * Its safe to read more than we are asked, caller should ensure that
	 * destination has enough space.
	 */
	src_paddr = round_down(src_paddr, 16);
	offset = src_paddr & 15;
	sz = round_up(sz + offset, 16);

	return __sev_issue_dbg_cmd(kvm, src_paddr, dst_paddr, sz, err, false);
}

static int __sev_dbg_decrypt_user(struct kvm *kvm, unsigned long paddr,
				  unsigned long __user dst_uaddr,
				  unsigned long dst_paddr,
				  int size, int *err)
{
	struct page *tpage = NULL;
	int ret, offset;

	/* if inputs are not 16-byte then use intermediate buffer */
	if (!IS_ALIGNED(dst_paddr, 16) ||
	    !IS_ALIGNED(paddr,     16) ||
	    !IS_ALIGNED(size,      16)) {
		tpage = (void *)alloc_page(GFP_KERNEL);
		if (!tpage)
			return -ENOMEM;

		dst_paddr = __sme_page_pa(tpage);
	}

	ret = __sev_dbg_decrypt(kvm, paddr, dst_paddr, size, err);
	if (ret)
		goto e_free;

	if (tpage) {
		offset = paddr & 15;
		if (copy_to_user((void __user *)(uintptr_t)dst_uaddr,
				 page_address(tpage) + offset, size))
			ret = -EFAULT;
	}

e_free:
	if (tpage)
		__free_page(tpage);

	return ret;
}

6481 6482 6483 6484 6485 6486 6487 6488 6489 6490 6491 6492 6493 6494 6495 6496 6497 6498 6499 6500 6501 6502 6503 6504 6505 6506 6507 6508 6509 6510 6511 6512 6513 6514 6515 6516 6517 6518 6519 6520 6521 6522 6523 6524 6525 6526 6527 6528 6529 6530 6531 6532 6533 6534 6535 6536 6537 6538 6539 6540 6541 6542 6543 6544 6545 6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556 6557
static int __sev_dbg_encrypt_user(struct kvm *kvm, unsigned long paddr,
				  unsigned long __user vaddr,
				  unsigned long dst_paddr,
				  unsigned long __user dst_vaddr,
				  int size, int *error)
{
	struct page *src_tpage = NULL;
	struct page *dst_tpage = NULL;
	int ret, len = size;

	/* If source buffer is not aligned then use an intermediate buffer */
	if (!IS_ALIGNED(vaddr, 16)) {
		src_tpage = alloc_page(GFP_KERNEL);
		if (!src_tpage)
			return -ENOMEM;

		if (copy_from_user(page_address(src_tpage),
				(void __user *)(uintptr_t)vaddr, size)) {
			__free_page(src_tpage);
			return -EFAULT;
		}

		paddr = __sme_page_pa(src_tpage);
	}

	/*
	 *  If destination buffer or length is not aligned then do read-modify-write:
	 *   - decrypt destination in an intermediate buffer
	 *   - copy the source buffer in an intermediate buffer
	 *   - use the intermediate buffer as source buffer
	 */
	if (!IS_ALIGNED(dst_vaddr, 16) || !IS_ALIGNED(size, 16)) {
		int dst_offset;

		dst_tpage = alloc_page(GFP_KERNEL);
		if (!dst_tpage) {
			ret = -ENOMEM;
			goto e_free;
		}

		ret = __sev_dbg_decrypt(kvm, dst_paddr,
					__sme_page_pa(dst_tpage), size, error);
		if (ret)
			goto e_free;

		/*
		 *  If source is kernel buffer then use memcpy() otherwise
		 *  copy_from_user().
		 */
		dst_offset = dst_paddr & 15;

		if (src_tpage)
			memcpy(page_address(dst_tpage) + dst_offset,
			       page_address(src_tpage), size);
		else {
			if (copy_from_user(page_address(dst_tpage) + dst_offset,
					   (void __user *)(uintptr_t)vaddr, size)) {
				ret = -EFAULT;
				goto e_free;
			}
		}

		paddr = __sme_page_pa(dst_tpage);
		dst_paddr = round_down(dst_paddr, 16);
		len = round_up(size, 16);
	}

	ret = __sev_issue_dbg_cmd(kvm, paddr, dst_paddr, len, error, true);

e_free:
	if (src_tpage)
		__free_page(src_tpage);
	if (dst_tpage)
		__free_page(dst_tpage);
	return ret;
}

6558 6559 6560 6561 6562 6563 6564 6565 6566 6567 6568 6569 6570 6571 6572 6573 6574 6575 6576 6577 6578 6579 6580 6581 6582 6583 6584 6585 6586 6587 6588 6589 6590 6591 6592 6593 6594 6595 6596 6597 6598 6599 6600 6601 6602 6603 6604 6605 6606 6607 6608 6609
static int sev_dbg_crypt(struct kvm *kvm, struct kvm_sev_cmd *argp, bool dec)
{
	unsigned long vaddr, vaddr_end, next_vaddr;
	unsigned long dst_vaddr, dst_vaddr_end;
	struct page **src_p, **dst_p;
	struct kvm_sev_dbg debug;
	unsigned long n;
	int ret, size;

	if (!sev_guest(kvm))
		return -ENOTTY;

	if (copy_from_user(&debug, (void __user *)(uintptr_t)argp->data, sizeof(debug)))
		return -EFAULT;

	vaddr = debug.src_uaddr;
	size = debug.len;
	vaddr_end = vaddr + size;
	dst_vaddr = debug.dst_uaddr;
	dst_vaddr_end = dst_vaddr + size;

	for (; vaddr < vaddr_end; vaddr = next_vaddr) {
		int len, s_off, d_off;

		/* lock userspace source and destination page */
		src_p = sev_pin_memory(kvm, vaddr & PAGE_MASK, PAGE_SIZE, &n, 0);
		if (!src_p)
			return -EFAULT;

		dst_p = sev_pin_memory(kvm, dst_vaddr & PAGE_MASK, PAGE_SIZE, &n, 1);
		if (!dst_p) {
			sev_unpin_memory(kvm, src_p, n);
			return -EFAULT;
		}

		/*
		 * The DBG_{DE,EN}CRYPT commands will perform {dec,en}cryption of the
		 * memory content (i.e it will write the same memory region with C=1).
		 * It's possible that the cache may contain the data with C=0, i.e.,
		 * unencrypted so invalidate it first.
		 */
		sev_clflush_pages(src_p, 1);
		sev_clflush_pages(dst_p, 1);

		/*
		 * Since user buffer may not be page aligned, calculate the
		 * offset within the page.
		 */
		s_off = vaddr & ~PAGE_MASK;
		d_off = dst_vaddr & ~PAGE_MASK;
		len = min_t(size_t, (PAGE_SIZE - s_off), size);

6610 6611 6612 6613 6614 6615 6616 6617 6618 6619 6620 6621 6622
		if (dec)
			ret = __sev_dbg_decrypt_user(kvm,
						     __sme_page_pa(src_p[0]) + s_off,
						     dst_vaddr,
						     __sme_page_pa(dst_p[0]) + d_off,
						     len, &argp->error);
		else
			ret = __sev_dbg_encrypt_user(kvm,
						     __sme_page_pa(src_p[0]) + s_off,
						     vaddr,
						     __sme_page_pa(dst_p[0]) + d_off,
						     dst_vaddr,
						     len, &argp->error);
6623 6624 6625 6626 6627 6628 6629 6630 6631 6632 6633 6634 6635 6636 6637

		sev_unpin_memory(kvm, src_p, 1);
		sev_unpin_memory(kvm, dst_p, 1);

		if (ret)
			goto err;

		next_vaddr = vaddr + len;
		dst_vaddr = dst_vaddr + len;
		size -= len;
	}
err:
	return ret;
}

6638 6639 6640 6641 6642 6643 6644 6645
static int sev_launch_secret(struct kvm *kvm, struct kvm_sev_cmd *argp)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct sev_data_launch_secret *data;
	struct kvm_sev_launch_secret params;
	struct page **pages;
	void *blob, *hdr;
	unsigned long n;
6646
	int ret, offset;
6647 6648 6649 6650 6651 6652 6653 6654 6655 6656 6657 6658 6659 6660 6661 6662 6663 6664 6665 6666 6667 6668 6669 6670 6671

	if (!sev_guest(kvm))
		return -ENOTTY;

	if (copy_from_user(&params, (void __user *)(uintptr_t)argp->data, sizeof(params)))
		return -EFAULT;

	pages = sev_pin_memory(kvm, params.guest_uaddr, params.guest_len, &n, 1);
	if (!pages)
		return -ENOMEM;

	/*
	 * The secret must be copied into contiguous memory region, lets verify
	 * that userspace memory pages are contiguous before we issue command.
	 */
	if (get_num_contig_pages(0, pages, n) != n) {
		ret = -EINVAL;
		goto e_unpin_memory;
	}

	ret = -ENOMEM;
	data = kzalloc(sizeof(*data), GFP_KERNEL);
	if (!data)
		goto e_unpin_memory;

6672 6673 6674 6675
	offset = params.guest_uaddr & (PAGE_SIZE - 1);
	data->guest_address = __sme_page_pa(pages[0]) + offset;
	data->guest_len = params.guest_len;

6676 6677 6678 6679 6680 6681 6682 6683 6684 6685 6686 6687 6688 6689
	blob = psp_copy_user_blob(params.trans_uaddr, params.trans_len);
	if (IS_ERR(blob)) {
		ret = PTR_ERR(blob);
		goto e_free;
	}

	data->trans_address = __psp_pa(blob);
	data->trans_len = params.trans_len;

	hdr = psp_copy_user_blob(params.hdr_uaddr, params.hdr_len);
	if (IS_ERR(hdr)) {
		ret = PTR_ERR(hdr);
		goto e_free_blob;
	}
6690 6691
	data->hdr_address = __psp_pa(hdr);
	data->hdr_len = params.hdr_len;
6692 6693 6694 6695 6696 6697 6698 6699 6700 6701 6702 6703 6704 6705 6706

	data->handle = sev->handle;
	ret = sev_issue_cmd(kvm, SEV_CMD_LAUNCH_UPDATE_SECRET, data, &argp->error);

	kfree(hdr);

e_free_blob:
	kfree(blob);
e_free:
	kfree(data);
e_unpin_memory:
	sev_unpin_memory(kvm, pages, n);
	return ret;
}

B
Brijesh Singh 已提交
6707 6708 6709 6710 6711 6712 6713 6714 6715 6716 6717 6718 6719 6720 6721 6722 6723
static int svm_mem_enc_op(struct kvm *kvm, void __user *argp)
{
	struct kvm_sev_cmd sev_cmd;
	int r;

	if (!svm_sev_enabled())
		return -ENOTTY;

	if (copy_from_user(&sev_cmd, argp, sizeof(struct kvm_sev_cmd)))
		return -EFAULT;

	mutex_lock(&kvm->lock);

	switch (sev_cmd.id) {
	case KVM_SEV_INIT:
		r = sev_guest_init(kvm, &sev_cmd);
		break;
6724 6725 6726
	case KVM_SEV_LAUNCH_START:
		r = sev_launch_start(kvm, &sev_cmd);
		break;
6727 6728 6729
	case KVM_SEV_LAUNCH_UPDATE_DATA:
		r = sev_launch_update_data(kvm, &sev_cmd);
		break;
6730 6731 6732
	case KVM_SEV_LAUNCH_MEASURE:
		r = sev_launch_measure(kvm, &sev_cmd);
		break;
6733 6734 6735
	case KVM_SEV_LAUNCH_FINISH:
		r = sev_launch_finish(kvm, &sev_cmd);
		break;
6736 6737 6738
	case KVM_SEV_GUEST_STATUS:
		r = sev_guest_status(kvm, &sev_cmd);
		break;
6739 6740 6741
	case KVM_SEV_DBG_DECRYPT:
		r = sev_dbg_crypt(kvm, &sev_cmd, true);
		break;
6742 6743 6744
	case KVM_SEV_DBG_ENCRYPT:
		r = sev_dbg_crypt(kvm, &sev_cmd, false);
		break;
6745 6746 6747
	case KVM_SEV_LAUNCH_SECRET:
		r = sev_launch_secret(kvm, &sev_cmd);
		break;
B
Brijesh Singh 已提交
6748 6749 6750 6751 6752 6753 6754 6755 6756 6757 6758 6759 6760
	default:
		r = -EINVAL;
		goto out;
	}

	if (copy_to_user(argp, &sev_cmd, sizeof(struct kvm_sev_cmd)))
		r = -EFAULT;

out:
	mutex_unlock(&kvm->lock);
	return r;
}

6761 6762 6763 6764 6765 6766 6767 6768 6769 6770 6771 6772 6773 6774 6775 6776 6777 6778 6779 6780 6781 6782 6783 6784 6785 6786 6787 6788 6789 6790 6791 6792 6793 6794 6795 6796 6797 6798 6799 6800 6801 6802 6803 6804 6805 6806 6807 6808 6809 6810 6811 6812 6813 6814 6815 6816 6817 6818 6819 6820 6821 6822 6823 6824 6825 6826 6827 6828 6829 6830 6831 6832 6833 6834 6835 6836 6837 6838 6839 6840 6841 6842 6843 6844 6845 6846 6847 6848
static int svm_register_enc_region(struct kvm *kvm,
				   struct kvm_enc_region *range)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct enc_region *region;
	int ret = 0;

	if (!sev_guest(kvm))
		return -ENOTTY;

	region = kzalloc(sizeof(*region), GFP_KERNEL);
	if (!region)
		return -ENOMEM;

	region->pages = sev_pin_memory(kvm, range->addr, range->size, &region->npages, 1);
	if (!region->pages) {
		ret = -ENOMEM;
		goto e_free;
	}

	/*
	 * The guest may change the memory encryption attribute from C=0 -> C=1
	 * or vice versa for this memory range. Lets make sure caches are
	 * flushed to ensure that guest data gets written into memory with
	 * correct C-bit.
	 */
	sev_clflush_pages(region->pages, region->npages);

	region->uaddr = range->addr;
	region->size = range->size;

	mutex_lock(&kvm->lock);
	list_add_tail(&region->list, &sev->regions_list);
	mutex_unlock(&kvm->lock);

	return ret;

e_free:
	kfree(region);
	return ret;
}

static struct enc_region *
find_enc_region(struct kvm *kvm, struct kvm_enc_region *range)
{
	struct kvm_sev_info *sev = &kvm->arch.sev_info;
	struct list_head *head = &sev->regions_list;
	struct enc_region *i;

	list_for_each_entry(i, head, list) {
		if (i->uaddr == range->addr &&
		    i->size == range->size)
			return i;
	}

	return NULL;
}


static int svm_unregister_enc_region(struct kvm *kvm,
				     struct kvm_enc_region *range)
{
	struct enc_region *region;
	int ret;

	mutex_lock(&kvm->lock);

	if (!sev_guest(kvm)) {
		ret = -ENOTTY;
		goto failed;
	}

	region = find_enc_region(kvm, range);
	if (!region) {
		ret = -EINVAL;
		goto failed;
	}

	__unregister_enc_region_locked(kvm, region);

	mutex_unlock(&kvm->lock);
	return 0;

failed:
	mutex_unlock(&kvm->lock);
	return ret;
}

6849
static struct kvm_x86_ops svm_x86_ops __ro_after_init = {
A
Avi Kivity 已提交
6850 6851 6852 6853
	.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
6854
	.check_processor_compatibility = svm_check_processor_compat,
A
Avi Kivity 已提交
6855 6856
	.hardware_enable = svm_hardware_enable,
	.hardware_disable = svm_hardware_disable,
6857
	.cpu_has_accelerated_tpr = svm_cpu_has_accelerated_tpr,
6858
	.cpu_has_high_real_mode_segbase = svm_has_high_real_mode_segbase,
A
Avi Kivity 已提交
6859 6860 6861

	.vcpu_create = svm_create_vcpu,
	.vcpu_free = svm_free_vcpu,
6862
	.vcpu_reset = svm_vcpu_reset,
A
Avi Kivity 已提交
6863

6864
	.vm_init = avic_vm_init,
B
Brijesh Singh 已提交
6865
	.vm_destroy = svm_vm_destroy,
6866

6867
	.prepare_guest_switch = svm_prepare_guest_switch,
A
Avi Kivity 已提交
6868 6869
	.vcpu_load = svm_vcpu_load,
	.vcpu_put = svm_vcpu_put,
6870 6871
	.vcpu_blocking = svm_vcpu_blocking,
	.vcpu_unblocking = svm_vcpu_unblocking,
A
Avi Kivity 已提交
6872

6873
	.update_bp_intercept = update_bp_intercept,
6874
	.get_msr_feature = svm_get_msr_feature,
A
Avi Kivity 已提交
6875 6876 6877 6878 6879
	.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,
6880
	.get_cpl = svm_get_cpl,
6881
	.get_cs_db_l_bits = kvm_get_cs_db_l_bits,
6882
	.decache_cr0_guest_bits = svm_decache_cr0_guest_bits,
6883
	.decache_cr3 = svm_decache_cr3,
6884
	.decache_cr4_guest_bits = svm_decache_cr4_guest_bits,
A
Avi Kivity 已提交
6885 6886 6887 6888 6889 6890 6891 6892
	.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
6893 6894
	.get_dr6 = svm_get_dr6,
	.set_dr6 = svm_set_dr6,
6895
	.set_dr7 = svm_set_dr7,
6896
	.sync_dirty_debug_regs = svm_sync_dirty_debug_regs,
A
Avi Kivity 已提交
6897
	.cache_reg = svm_cache_reg,
A
Avi Kivity 已提交
6898 6899
	.get_rflags = svm_get_rflags,
	.set_rflags = svm_set_rflags,
6900

A
Avi Kivity 已提交
6901 6902 6903
	.tlb_flush = svm_flush_tlb,

	.run = svm_vcpu_run,
6904
	.handle_exit = handle_exit,
A
Avi Kivity 已提交
6905
	.skip_emulated_instruction = skip_emulated_instruction,
6906 6907
	.set_interrupt_shadow = svm_set_interrupt_shadow,
	.get_interrupt_shadow = svm_get_interrupt_shadow,
I
Ingo Molnar 已提交
6908
	.patch_hypercall = svm_patch_hypercall,
E
Eddie Dong 已提交
6909
	.set_irq = svm_set_irq,
6910
	.set_nmi = svm_inject_nmi,
6911
	.queue_exception = svm_queue_exception,
A
Avi Kivity 已提交
6912
	.cancel_injection = svm_cancel_injection,
6913
	.interrupt_allowed = svm_interrupt_allowed,
6914
	.nmi_allowed = svm_nmi_allowed,
J
Jan Kiszka 已提交
6915 6916
	.get_nmi_mask = svm_get_nmi_mask,
	.set_nmi_mask = svm_set_nmi_mask,
6917 6918 6919
	.enable_nmi_window = enable_nmi_window,
	.enable_irq_window = enable_irq_window,
	.update_cr8_intercept = update_cr8_intercept,
6920
	.set_virtual_x2apic_mode = svm_set_virtual_x2apic_mode,
6921 6922
	.get_enable_apicv = svm_get_enable_apicv,
	.refresh_apicv_exec_ctrl = svm_refresh_apicv_exec_ctrl,
6923
	.load_eoi_exitmap = svm_load_eoi_exitmap,
6924 6925
	.hwapic_irr_update = svm_hwapic_irr_update,
	.hwapic_isr_update = svm_hwapic_isr_update,
6926
	.sync_pir_to_irr = kvm_lapic_find_highest_irr,
6927
	.apicv_post_state_restore = avic_post_state_restore,
6928 6929

	.set_tss_addr = svm_set_tss_addr,
6930
	.get_tdp_level = get_npt_level,
6931
	.get_mt_mask = svm_get_mt_mask,
6932

6933 6934
	.get_exit_info = svm_get_exit_info,

6935
	.get_lpage_level = svm_get_lpage_level,
6936 6937

	.cpuid_update = svm_cpuid_update,
6938 6939

	.rdtscp_supported = svm_rdtscp_supported,
6940
	.invpcid_supported = svm_invpcid_supported,
6941
	.mpx_supported = svm_mpx_supported,
6942
	.xsaves_supported = svm_xsaves_supported,
6943
	.umip_emulated = svm_umip_emulated,
6944 6945

	.set_supported_cpuid = svm_set_supported_cpuid,
6946 6947

	.has_wbinvd_exit = svm_has_wbinvd_exit,
6948 6949

	.write_tsc_offset = svm_write_tsc_offset,
6950 6951

	.set_tdp_cr3 = set_tdp_cr3,
6952 6953

	.check_intercept = svm_check_intercept,
6954
	.handle_external_intr = svm_handle_external_intr,
6955 6956

	.sched_in = svm_sched_in,
6957 6958

	.pmu_ops = &amd_pmu_ops,
6959
	.deliver_posted_interrupt = svm_deliver_avic_intr,
6960
	.update_pi_irte = svm_update_pi_irte,
6961
	.setup_mce = svm_setup_mce,
6962

6963
	.smi_allowed = svm_smi_allowed,
6964 6965
	.pre_enter_smm = svm_pre_enter_smm,
	.pre_leave_smm = svm_pre_leave_smm,
6966
	.enable_smi_window = enable_smi_window,
B
Brijesh Singh 已提交
6967 6968

	.mem_enc_op = svm_mem_enc_op,
6969 6970
	.mem_enc_reg_region = svm_register_enc_region,
	.mem_enc_unreg_region = svm_unregister_enc_region,
A
Avi Kivity 已提交
6971 6972 6973 6974
};

static int __init svm_init(void)
{
6975
	return kvm_init(&svm_x86_ops, sizeof(struct vcpu_svm),
6976
			__alignof__(struct vcpu_svm), THIS_MODULE);
A
Avi Kivity 已提交
6977 6978 6979 6980
}

static void __exit svm_exit(void)
{
6981
	kvm_exit();
A
Avi Kivity 已提交
6982 6983 6984 6985
}

module_init(svm_init)
module_exit(svm_exit)