x86.c 94.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
/*
 * Kernel-based Virtual Machine driver for Linux
 *
 * derived from drivers/kvm/kvm_main.c
 *
 * Copyright (C) 2006 Qumranet, Inc.
 *
 * Authors:
 *   Avi Kivity   <avi@qumranet.com>
 *   Yaniv Kamay  <yaniv@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
#include <linux/kvm_host.h>
18
#include "irq.h"
19
#include "mmu.h"
S
Sheng Yang 已提交
20
#include "i8254.h"
21
#include "tss.h"
22

23
#include <linux/clocksource.h>
24 25 26
#include <linux/kvm.h>
#include <linux/fs.h>
#include <linux/vmalloc.h>
27
#include <linux/module.h>
28
#include <linux/mman.h>
29
#include <linux/highmem.h>
30 31

#include <asm/uaccess.h>
32
#include <asm/msr.h>
33
#include <asm/desc.h>
34

35
#define MAX_IO_MSRS 256
36 37 38 39 40 41 42 43 44 45 46
#define CR0_RESERVED_BITS						\
	(~(unsigned long)(X86_CR0_PE | X86_CR0_MP | X86_CR0_EM | X86_CR0_TS \
			  | X86_CR0_ET | X86_CR0_NE | X86_CR0_WP | X86_CR0_AM \
			  | X86_CR0_NW | X86_CR0_CD | X86_CR0_PG))
#define CR4_RESERVED_BITS						\
	(~(unsigned long)(X86_CR4_VME | X86_CR4_PVI | X86_CR4_TSD | X86_CR4_DE\
			  | X86_CR4_PSE | X86_CR4_PAE | X86_CR4_MCE	\
			  | X86_CR4_PGE | X86_CR4_PCE | X86_CR4_OSFXSR	\
			  | X86_CR4_OSXMMEXCPT | X86_CR4_VMXE))

#define CR8_RESERVED_BITS (~(unsigned long)X86_CR8_TPR)
47 48 49 50 51 52 53 54 55
/* EFER defaults:
 * - enable syscall per default because its emulated by KVM
 * - enable LME and LMA per default on 64 bit KVM
 */
#ifdef CONFIG_X86_64
static u64 __read_mostly efer_reserved_bits = 0xfffffffffffffafeULL;
#else
static u64 __read_mostly efer_reserved_bits = 0xfffffffffffffffeULL;
#endif
56

57 58
#define VM_STAT(x) offsetof(struct kvm, stat.x), KVM_STAT_VM
#define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU
59

60 61 62
static int kvm_dev_ioctl_get_supported_cpuid(struct kvm_cpuid2 *cpuid,
				    struct kvm_cpuid_entry2 __user *entries);

63 64
struct kvm_x86_ops *kvm_x86_ops;

65
struct kvm_stats_debugfs_item debugfs_entries[] = {
66 67 68 69 70 71 72 73 74 75 76
	{ "pf_fixed", VCPU_STAT(pf_fixed) },
	{ "pf_guest", VCPU_STAT(pf_guest) },
	{ "tlb_flush", VCPU_STAT(tlb_flush) },
	{ "invlpg", VCPU_STAT(invlpg) },
	{ "exits", VCPU_STAT(exits) },
	{ "io_exits", VCPU_STAT(io_exits) },
	{ "mmio_exits", VCPU_STAT(mmio_exits) },
	{ "signal_exits", VCPU_STAT(signal_exits) },
	{ "irq_window", VCPU_STAT(irq_window_exits) },
	{ "halt_exits", VCPU_STAT(halt_exits) },
	{ "halt_wakeup", VCPU_STAT(halt_wakeup) },
A
Amit Shah 已提交
77
	{ "hypercalls", VCPU_STAT(hypercalls) },
78 79 80 81 82 83 84
	{ "request_irq", VCPU_STAT(request_irq_exits) },
	{ "irq_exits", VCPU_STAT(irq_exits) },
	{ "host_state_reload", VCPU_STAT(host_state_reload) },
	{ "efer_reload", VCPU_STAT(efer_reload) },
	{ "fpu_reload", VCPU_STAT(fpu_reload) },
	{ "insn_emulation", VCPU_STAT(insn_emulation) },
	{ "insn_emulation_fail", VCPU_STAT(insn_emulation_fail) },
A
Avi Kivity 已提交
85 86 87 88 89 90
	{ "mmu_shadow_zapped", VM_STAT(mmu_shadow_zapped) },
	{ "mmu_pte_write", VM_STAT(mmu_pte_write) },
	{ "mmu_pte_updated", VM_STAT(mmu_pte_updated) },
	{ "mmu_pde_zapped", VM_STAT(mmu_pde_zapped) },
	{ "mmu_flooded", VM_STAT(mmu_flooded) },
	{ "mmu_recycled", VM_STAT(mmu_recycled) },
A
Avi Kivity 已提交
91
	{ "mmu_cache_miss", VM_STAT(mmu_cache_miss) },
92
	{ "remote_tlb_flush", VM_STAT(remote_tlb_flush) },
M
Marcelo Tosatti 已提交
93
	{ "largepages", VM_STAT(lpages) },
94 95 96 97
	{ NULL }
};


98 99 100
unsigned long segment_base(u16 selector)
{
	struct descriptor_table gdt;
101
	struct desc_struct *d;
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
	unsigned long table_base;
	unsigned long v;

	if (selector == 0)
		return 0;

	asm("sgdt %0" : "=m"(gdt));
	table_base = gdt.base;

	if (selector & 4) {           /* from ldt */
		u16 ldt_selector;

		asm("sldt %0" : "=g"(ldt_selector));
		table_base = segment_base(ldt_selector);
	}
117 118 119
	d = (struct desc_struct *)(table_base + (selector & ~7));
	v = d->base0 | ((unsigned long)d->base1 << 16) |
		((unsigned long)d->base2 << 24);
120
#ifdef CONFIG_X86_64
121 122
	if (d->s == 0 && (d->type == 2 || d->type == 9 || d->type == 11))
		v |= ((unsigned long)((struct ldttss_desc64 *)d)->base3) << 32;
123 124 125 126 127
#endif
	return v;
}
EXPORT_SYMBOL_GPL(segment_base);

128 129 130
u64 kvm_get_apic_base(struct kvm_vcpu *vcpu)
{
	if (irqchip_in_kernel(vcpu->kvm))
131
		return vcpu->arch.apic_base;
132
	else
133
		return vcpu->arch.apic_base;
134 135 136 137 138 139 140 141 142
}
EXPORT_SYMBOL_GPL(kvm_get_apic_base);

void kvm_set_apic_base(struct kvm_vcpu *vcpu, u64 data)
{
	/* TODO: reserve bits check */
	if (irqchip_in_kernel(vcpu->kvm))
		kvm_lapic_set_base(vcpu, data);
	else
143
		vcpu->arch.apic_base = data;
144 145 146
}
EXPORT_SYMBOL_GPL(kvm_set_apic_base);

147 148
void kvm_queue_exception(struct kvm_vcpu *vcpu, unsigned nr)
{
149 150 151 152
	WARN_ON(vcpu->arch.exception.pending);
	vcpu->arch.exception.pending = true;
	vcpu->arch.exception.has_error_code = false;
	vcpu->arch.exception.nr = nr;
153 154 155
}
EXPORT_SYMBOL_GPL(kvm_queue_exception);

156 157 158 159
void kvm_inject_page_fault(struct kvm_vcpu *vcpu, unsigned long addr,
			   u32 error_code)
{
	++vcpu->stat.pf_guest;
J
Joerg Roedel 已提交
160 161 162 163 164 165 166 167 168 169
	if (vcpu->arch.exception.pending) {
		if (vcpu->arch.exception.nr == PF_VECTOR) {
			printk(KERN_DEBUG "kvm: inject_page_fault:"
					" double fault 0x%lx\n", addr);
			vcpu->arch.exception.nr = DF_VECTOR;
			vcpu->arch.exception.error_code = 0;
		} else if (vcpu->arch.exception.nr == DF_VECTOR) {
			/* triple fault -> shutdown */
			set_bit(KVM_REQ_TRIPLE_FAULT, &vcpu->requests);
		}
170 171
		return;
	}
172
	vcpu->arch.cr2 = addr;
173 174 175
	kvm_queue_exception_e(vcpu, PF_VECTOR, error_code);
}

176 177
void kvm_queue_exception_e(struct kvm_vcpu *vcpu, unsigned nr, u32 error_code)
{
178 179 180 181 182
	WARN_ON(vcpu->arch.exception.pending);
	vcpu->arch.exception.pending = true;
	vcpu->arch.exception.has_error_code = true;
	vcpu->arch.exception.nr = nr;
	vcpu->arch.exception.error_code = error_code;
183 184 185 186 187
}
EXPORT_SYMBOL_GPL(kvm_queue_exception_e);

static void __queue_exception(struct kvm_vcpu *vcpu)
{
188 189 190
	kvm_x86_ops->queue_exception(vcpu, vcpu->arch.exception.nr,
				     vcpu->arch.exception.has_error_code,
				     vcpu->arch.exception.error_code);
191 192
}

193 194 195 196 197 198 199 200 201
/*
 * Load the pae pdptrs.  Return true is they are all valid.
 */
int load_pdptrs(struct kvm_vcpu *vcpu, unsigned long cr3)
{
	gfn_t pdpt_gfn = cr3 >> PAGE_SHIFT;
	unsigned offset = ((cr3 & (PAGE_SIZE-1)) >> 5) << 2;
	int i;
	int ret;
202
	u64 pdpte[ARRAY_SIZE(vcpu->arch.pdptrs)];
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217

	ret = kvm_read_guest_page(vcpu->kvm, pdpt_gfn, pdpte,
				  offset * sizeof(u64), sizeof(pdpte));
	if (ret < 0) {
		ret = 0;
		goto out;
	}
	for (i = 0; i < ARRAY_SIZE(pdpte); ++i) {
		if ((pdpte[i] & 1) && (pdpte[i] & 0xfffffff0000001e6ull)) {
			ret = 0;
			goto out;
		}
	}
	ret = 1;

218
	memcpy(vcpu->arch.pdptrs, pdpte, sizeof(vcpu->arch.pdptrs));
219 220 221 222
out:

	return ret;
}
223
EXPORT_SYMBOL_GPL(load_pdptrs);
224

225 226
static bool pdptrs_changed(struct kvm_vcpu *vcpu)
{
227
	u64 pdpte[ARRAY_SIZE(vcpu->arch.pdptrs)];
228 229 230 231 232 233
	bool changed = true;
	int r;

	if (is_long_mode(vcpu) || !is_pae(vcpu))
		return false;

234
	r = kvm_read_guest(vcpu->kvm, vcpu->arch.cr3 & ~31u, pdpte, sizeof(pdpte));
235 236
	if (r < 0)
		goto out;
237
	changed = memcmp(pdpte, vcpu->arch.pdptrs, sizeof(pdpte)) != 0;
238 239 240 241 242
out:

	return changed;
}

243
void kvm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
244 245 246
{
	if (cr0 & CR0_RESERVED_BITS) {
		printk(KERN_DEBUG "set_cr0: 0x%lx #GP, reserved bits 0x%lx\n",
247
		       cr0, vcpu->arch.cr0);
248
		kvm_inject_gp(vcpu, 0);
249 250 251 252 253
		return;
	}

	if ((cr0 & X86_CR0_NW) && !(cr0 & X86_CR0_CD)) {
		printk(KERN_DEBUG "set_cr0: #GP, CD == 0 && NW == 1\n");
254
		kvm_inject_gp(vcpu, 0);
255 256 257 258 259 260
		return;
	}

	if ((cr0 & X86_CR0_PG) && !(cr0 & X86_CR0_PE)) {
		printk(KERN_DEBUG "set_cr0: #GP, set PG flag "
		       "and a clear PE flag\n");
261
		kvm_inject_gp(vcpu, 0);
262 263 264 265 266
		return;
	}

	if (!is_paging(vcpu) && (cr0 & X86_CR0_PG)) {
#ifdef CONFIG_X86_64
267
		if ((vcpu->arch.shadow_efer & EFER_LME)) {
268 269 270 271 272
			int cs_db, cs_l;

			if (!is_pae(vcpu)) {
				printk(KERN_DEBUG "set_cr0: #GP, start paging "
				       "in long mode while PAE is disabled\n");
273
				kvm_inject_gp(vcpu, 0);
274 275 276 277 278 279
				return;
			}
			kvm_x86_ops->get_cs_db_l_bits(vcpu, &cs_db, &cs_l);
			if (cs_l) {
				printk(KERN_DEBUG "set_cr0: #GP, start paging "
				       "in long mode while CS.L == 1\n");
280
				kvm_inject_gp(vcpu, 0);
281 282 283 284 285
				return;

			}
		} else
#endif
286
		if (is_pae(vcpu) && !load_pdptrs(vcpu, vcpu->arch.cr3)) {
287 288
			printk(KERN_DEBUG "set_cr0: #GP, pdptrs "
			       "reserved bits\n");
289
			kvm_inject_gp(vcpu, 0);
290 291 292 293 294 295
			return;
		}

	}

	kvm_x86_ops->set_cr0(vcpu, cr0);
296
	vcpu->arch.cr0 = cr0;
297 298 299 300

	kvm_mmu_reset_context(vcpu);
	return;
}
301
EXPORT_SYMBOL_GPL(kvm_set_cr0);
302

303
void kvm_lmsw(struct kvm_vcpu *vcpu, unsigned long msw)
304
{
305
	kvm_set_cr0(vcpu, (vcpu->arch.cr0 & ~0x0ful) | (msw & 0x0f));
F
Feng (Eric) Liu 已提交
306 307 308
	KVMTRACE_1D(LMSW, vcpu,
		    (u32)((vcpu->arch.cr0 & ~0x0ful) | (msw & 0x0f)),
		    handler);
309
}
310
EXPORT_SYMBOL_GPL(kvm_lmsw);
311

312
void kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
313 314 315
{
	if (cr4 & CR4_RESERVED_BITS) {
		printk(KERN_DEBUG "set_cr4: #GP, reserved bits\n");
316
		kvm_inject_gp(vcpu, 0);
317 318 319 320 321 322 323
		return;
	}

	if (is_long_mode(vcpu)) {
		if (!(cr4 & X86_CR4_PAE)) {
			printk(KERN_DEBUG "set_cr4: #GP, clearing PAE while "
			       "in long mode\n");
324
			kvm_inject_gp(vcpu, 0);
325 326 327
			return;
		}
	} else if (is_paging(vcpu) && !is_pae(vcpu) && (cr4 & X86_CR4_PAE)
328
		   && !load_pdptrs(vcpu, vcpu->arch.cr3)) {
329
		printk(KERN_DEBUG "set_cr4: #GP, pdptrs reserved bits\n");
330
		kvm_inject_gp(vcpu, 0);
331 332 333 334 335
		return;
	}

	if (cr4 & X86_CR4_VMXE) {
		printk(KERN_DEBUG "set_cr4: #GP, setting VMXE\n");
336
		kvm_inject_gp(vcpu, 0);
337 338 339
		return;
	}
	kvm_x86_ops->set_cr4(vcpu, cr4);
340
	vcpu->arch.cr4 = cr4;
341 342
	kvm_mmu_reset_context(vcpu);
}
343
EXPORT_SYMBOL_GPL(kvm_set_cr4);
344

345
void kvm_set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)
346
{
347
	if (cr3 == vcpu->arch.cr3 && !pdptrs_changed(vcpu)) {
348 349 350 351
		kvm_mmu_flush_tlb(vcpu);
		return;
	}

352 353 354
	if (is_long_mode(vcpu)) {
		if (cr3 & CR3_L_MODE_RESERVED_BITS) {
			printk(KERN_DEBUG "set_cr3: #GP, reserved bits\n");
355
			kvm_inject_gp(vcpu, 0);
356 357 358 359 360 361 362
			return;
		}
	} else {
		if (is_pae(vcpu)) {
			if (cr3 & CR3_PAE_RESERVED_BITS) {
				printk(KERN_DEBUG
				       "set_cr3: #GP, reserved bits\n");
363
				kvm_inject_gp(vcpu, 0);
364 365 366 367 368
				return;
			}
			if (is_paging(vcpu) && !load_pdptrs(vcpu, cr3)) {
				printk(KERN_DEBUG "set_cr3: #GP, pdptrs "
				       "reserved bits\n");
369
				kvm_inject_gp(vcpu, 0);
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388
				return;
			}
		}
		/*
		 * We don't check reserved bits in nonpae mode, because
		 * this isn't enforced, and VMware depends on this.
		 */
	}

	/*
	 * Does the new cr3 value map to physical memory? (Note, we
	 * catch an invalid cr3 even in real-mode, because it would
	 * cause trouble later on when we turn on paging anyway.)
	 *
	 * A real CPU would silently accept an invalid cr3 and would
	 * attempt to use it - with largely undefined (and often hard
	 * to debug) behavior on the guest side.
	 */
	if (unlikely(!gfn_to_memslot(vcpu->kvm, cr3 >> PAGE_SHIFT)))
389
		kvm_inject_gp(vcpu, 0);
390
	else {
391 392
		vcpu->arch.cr3 = cr3;
		vcpu->arch.mmu.new_cr3(vcpu);
393 394
	}
}
395
EXPORT_SYMBOL_GPL(kvm_set_cr3);
396

397
void kvm_set_cr8(struct kvm_vcpu *vcpu, unsigned long cr8)
398 399 400
{
	if (cr8 & CR8_RESERVED_BITS) {
		printk(KERN_DEBUG "set_cr8: #GP, reserved bits 0x%lx\n", cr8);
401
		kvm_inject_gp(vcpu, 0);
402 403 404 405 406
		return;
	}
	if (irqchip_in_kernel(vcpu->kvm))
		kvm_lapic_set_tpr(vcpu, cr8);
	else
407
		vcpu->arch.cr8 = cr8;
408
}
409
EXPORT_SYMBOL_GPL(kvm_set_cr8);
410

411
unsigned long kvm_get_cr8(struct kvm_vcpu *vcpu)
412 413 414 415
{
	if (irqchip_in_kernel(vcpu->kvm))
		return kvm_lapic_get_cr8(vcpu);
	else
416
		return vcpu->arch.cr8;
417
}
418
EXPORT_SYMBOL_GPL(kvm_get_cr8);
419

420 421 422 423 424 425 426 427 428 429 430 431 432
/*
 * List of msr numbers which we expose to userspace through KVM_GET_MSRS
 * and KVM_SET_MSRS, and KVM_GET_MSR_INDEX_LIST.
 *
 * This list is modified at module load time to reflect the
 * capabilities of the host cpu.
 */
static u32 msrs_to_save[] = {
	MSR_IA32_SYSENTER_CS, MSR_IA32_SYSENTER_ESP, MSR_IA32_SYSENTER_EIP,
	MSR_K6_STAR,
#ifdef CONFIG_X86_64
	MSR_CSTAR, MSR_KERNEL_GS_BASE, MSR_SYSCALL_MASK, MSR_LSTAR,
#endif
433
	MSR_IA32_TIME_STAMP_COUNTER, MSR_KVM_SYSTEM_TIME, MSR_KVM_WALL_CLOCK,
434
	MSR_IA32_PERF_STATUS,
435 436 437 438 439 440 441 442
};

static unsigned num_msrs_to_save;

static u32 emulated_msrs[] = {
	MSR_IA32_MISC_ENABLE,
};

443 444
static void set_efer(struct kvm_vcpu *vcpu, u64 efer)
{
445
	if (efer & efer_reserved_bits) {
446 447
		printk(KERN_DEBUG "set_efer: 0x%llx #GP, reserved bits\n",
		       efer);
448
		kvm_inject_gp(vcpu, 0);
449 450 451 452
		return;
	}

	if (is_paging(vcpu)
453
	    && (vcpu->arch.shadow_efer & EFER_LME) != (efer & EFER_LME)) {
454
		printk(KERN_DEBUG "set_efer: #GP, change LME while paging\n");
455
		kvm_inject_gp(vcpu, 0);
456 457 458 459 460 461
		return;
	}

	kvm_x86_ops->set_efer(vcpu, efer);

	efer &= ~EFER_LMA;
462
	efer |= vcpu->arch.shadow_efer & EFER_LMA;
463

464
	vcpu->arch.shadow_efer = efer;
465 466
}

467 468 469 470 471 472 473
void kvm_enable_efer_bits(u64 mask)
{
       efer_reserved_bits &= ~mask;
}
EXPORT_SYMBOL_GPL(kvm_enable_efer_bits);


474 475 476 477 478 479 480 481 482 483
/*
 * Writes msr value into into the appropriate "register".
 * Returns 0 on success, non-0 otherwise.
 * Assumes vcpu_load() was already called.
 */
int kvm_set_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data)
{
	return kvm_x86_ops->set_msr(vcpu, msr_index, data);
}

484 485 486 487 488 489 490 491
/*
 * Adapt set_msr() to msr_io()'s calling convention
 */
static int do_set_msr(struct kvm_vcpu *vcpu, unsigned index, u64 *data)
{
	return kvm_set_msr(vcpu, index, *data);
}

492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553
static void kvm_write_wall_clock(struct kvm *kvm, gpa_t wall_clock)
{
	static int version;
	struct kvm_wall_clock wc;
	struct timespec wc_ts;

	if (!wall_clock)
		return;

	version++;

	kvm_write_guest(kvm, wall_clock, &version, sizeof(version));

	wc_ts = current_kernel_time();
	wc.wc_sec = wc_ts.tv_sec;
	wc.wc_nsec = wc_ts.tv_nsec;
	wc.wc_version = version;

	kvm_write_guest(kvm, wall_clock, &wc, sizeof(wc));

	version++;
	kvm_write_guest(kvm, wall_clock, &version, sizeof(version));
}

static void kvm_write_guest_time(struct kvm_vcpu *v)
{
	struct timespec ts;
	unsigned long flags;
	struct kvm_vcpu_arch *vcpu = &v->arch;
	void *shared_kaddr;

	if ((!vcpu->time_page))
		return;

	/* Keep irq disabled to prevent changes to the clock */
	local_irq_save(flags);
	kvm_get_msr(v, MSR_IA32_TIME_STAMP_COUNTER,
			  &vcpu->hv_clock.tsc_timestamp);
	ktime_get_ts(&ts);
	local_irq_restore(flags);

	/* With all the info we got, fill in the values */

	vcpu->hv_clock.system_time = ts.tv_nsec +
				     (NSEC_PER_SEC * (u64)ts.tv_sec);
	/*
	 * The interface expects us to write an even number signaling that the
	 * update is finished. Since the guest won't see the intermediate
	 * state, we just write "2" at the end
	 */
	vcpu->hv_clock.version = 2;

	shared_kaddr = kmap_atomic(vcpu->time_page, KM_USER0);

	memcpy(shared_kaddr + vcpu->time_offset, &vcpu->hv_clock,
		sizeof(vcpu->hv_clock));

	kunmap_atomic(shared_kaddr, KM_USER0);

	mark_page_dirty(v->kvm, vcpu->time >> PAGE_SHIFT);
}

554 555 556 557 558 559 560 561 562

int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
{
	switch (msr) {
	case MSR_EFER:
		set_efer(vcpu, data);
		break;
	case MSR_IA32_MC0_STATUS:
		pr_unimpl(vcpu, "%s: MSR_IA32_MC0_STATUS 0x%llx, nop\n",
563
		       __func__, data);
564 565 566
		break;
	case MSR_IA32_MCG_STATUS:
		pr_unimpl(vcpu, "%s: MSR_IA32_MCG_STATUS 0x%llx, nop\n",
567
			__func__, data);
568
		break;
569 570
	case MSR_IA32_MCG_CTL:
		pr_unimpl(vcpu, "%s: MSR_IA32_MCG_CTL 0x%llx, nop\n",
571
			__func__, data);
572
		break;
573 574 575 576 577 578 579 580
	case MSR_IA32_UCODE_REV:
	case MSR_IA32_UCODE_WRITE:
	case 0x200 ... 0x2ff: /* MTRRs */
		break;
	case MSR_IA32_APICBASE:
		kvm_set_apic_base(vcpu, data);
		break;
	case MSR_IA32_MISC_ENABLE:
581
		vcpu->arch.ia32_misc_enable_msr = data;
582
		break;
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618
	case MSR_KVM_WALL_CLOCK:
		vcpu->kvm->arch.wall_clock = data;
		kvm_write_wall_clock(vcpu->kvm, data);
		break;
	case MSR_KVM_SYSTEM_TIME: {
		if (vcpu->arch.time_page) {
			kvm_release_page_dirty(vcpu->arch.time_page);
			vcpu->arch.time_page = NULL;
		}

		vcpu->arch.time = data;

		/* we verify if the enable bit is set... */
		if (!(data & 1))
			break;

		/* ...but clean it before doing the actual write */
		vcpu->arch.time_offset = data & ~(PAGE_MASK | 1);

		vcpu->arch.hv_clock.tsc_to_system_mul =
					clocksource_khz2mult(tsc_khz, 22);
		vcpu->arch.hv_clock.tsc_shift = 22;

		down_read(&current->mm->mmap_sem);
		vcpu->arch.time_page =
				gfn_to_page(vcpu->kvm, data >> PAGE_SHIFT);
		up_read(&current->mm->mmap_sem);

		if (is_error_page(vcpu->arch.time_page)) {
			kvm_release_page_clean(vcpu->arch.time_page);
			vcpu->arch.time_page = NULL;
		}

		kvm_write_guest_time(vcpu);
		break;
	}
619
	default:
620
		pr_unimpl(vcpu, "unhandled wrmsr: 0x%x data %llx\n", msr, data);
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650
		return 1;
	}
	return 0;
}
EXPORT_SYMBOL_GPL(kvm_set_msr_common);


/*
 * Reads an msr value (of 'msr_index') into 'pdata'.
 * Returns 0 on success, non-0 otherwise.
 * Assumes vcpu_load() was already called.
 */
int kvm_get_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 *pdata)
{
	return kvm_x86_ops->get_msr(vcpu, msr_index, pdata);
}

int kvm_get_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
{
	u64 data;

	switch (msr) {
	case 0xc0010010: /* SYSCFG */
	case 0xc0010015: /* HWCR */
	case MSR_IA32_PLATFORM_ID:
	case MSR_IA32_P5_MC_ADDR:
	case MSR_IA32_P5_MC_TYPE:
	case MSR_IA32_MC0_CTL:
	case MSR_IA32_MCG_STATUS:
	case MSR_IA32_MCG_CAP:
651
	case MSR_IA32_MCG_CTL:
652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670
	case MSR_IA32_MC0_MISC:
	case MSR_IA32_MC0_MISC+4:
	case MSR_IA32_MC0_MISC+8:
	case MSR_IA32_MC0_MISC+12:
	case MSR_IA32_MC0_MISC+16:
	case MSR_IA32_UCODE_REV:
	case MSR_IA32_EBL_CR_POWERON:
		/* MTRR registers */
	case 0xfe:
	case 0x200 ... 0x2ff:
		data = 0;
		break;
	case 0xcd: /* fsb frequency */
		data = 3;
		break;
	case MSR_IA32_APICBASE:
		data = kvm_get_apic_base(vcpu);
		break;
	case MSR_IA32_MISC_ENABLE:
671
		data = vcpu->arch.ia32_misc_enable_msr;
672
		break;
673 674 675 676 677 678
	case MSR_IA32_PERF_STATUS:
		/* TSC increment by tick */
		data = 1000ULL;
		/* CPU multiplier */
		data |= (((uint64_t)4ULL) << 40);
		break;
679
	case MSR_EFER:
680
		data = vcpu->arch.shadow_efer;
681
		break;
682 683 684 685 686 687
	case MSR_KVM_WALL_CLOCK:
		data = vcpu->kvm->arch.wall_clock;
		break;
	case MSR_KVM_SYSTEM_TIME:
		data = vcpu->arch.time;
		break;
688 689 690 691 692 693 694 695 696
	default:
		pr_unimpl(vcpu, "unhandled rdmsr: 0x%x\n", msr);
		return 1;
	}
	*pdata = data;
	return 0;
}
EXPORT_SYMBOL_GPL(kvm_get_msr_common);

697 698 699 700 701 702 703 704 705 706 707 708 709 710
/*
 * Read or write a bunch of msrs. All parameters are kernel addresses.
 *
 * @return number of msrs set successfully.
 */
static int __msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs *msrs,
		    struct kvm_msr_entry *entries,
		    int (*do_msr)(struct kvm_vcpu *vcpu,
				  unsigned index, u64 *data))
{
	int i;

	vcpu_load(vcpu);

711
	down_read(&vcpu->kvm->slots_lock);
712 713 714
	for (i = 0; i < msrs->nmsrs; ++i)
		if (do_msr(vcpu, entries[i].index, &entries[i].data))
			break;
715
	up_read(&vcpu->kvm->slots_lock);
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770

	vcpu_put(vcpu);

	return i;
}

/*
 * Read or write a bunch of msrs. Parameters are user addresses.
 *
 * @return number of msrs set successfully.
 */
static int msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs __user *user_msrs,
		  int (*do_msr)(struct kvm_vcpu *vcpu,
				unsigned index, u64 *data),
		  int writeback)
{
	struct kvm_msrs msrs;
	struct kvm_msr_entry *entries;
	int r, n;
	unsigned size;

	r = -EFAULT;
	if (copy_from_user(&msrs, user_msrs, sizeof msrs))
		goto out;

	r = -E2BIG;
	if (msrs.nmsrs >= MAX_IO_MSRS)
		goto out;

	r = -ENOMEM;
	size = sizeof(struct kvm_msr_entry) * msrs.nmsrs;
	entries = vmalloc(size);
	if (!entries)
		goto out;

	r = -EFAULT;
	if (copy_from_user(entries, user_msrs->entries, size))
		goto out_free;

	r = n = __msr_io(vcpu, &msrs, entries, do_msr);
	if (r < 0)
		goto out_free;

	r = -EFAULT;
	if (writeback && copy_to_user(user_msrs->entries, entries, size))
		goto out_free;

	r = n;

out_free:
	vfree(entries);
out:
	return r;
}

771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805
/*
 * Make sure that a cpu that is being hot-unplugged does not have any vcpus
 * cached on it.
 */
void decache_vcpus_on_cpu(int cpu)
{
	struct kvm *vm;
	struct kvm_vcpu *vcpu;
	int i;

	spin_lock(&kvm_lock);
	list_for_each_entry(vm, &vm_list, vm_list)
		for (i = 0; i < KVM_MAX_VCPUS; ++i) {
			vcpu = vm->vcpus[i];
			if (!vcpu)
				continue;
			/*
			 * If the vcpu is locked, then it is running on some
			 * other cpu and therefore it is not cached on the
			 * cpu in question.
			 *
			 * If it's not locked, check the last cpu it executed
			 * on.
			 */
			if (mutex_trylock(&vcpu->mutex)) {
				if (vcpu->cpu == cpu) {
					kvm_x86_ops->vcpu_decache(vcpu);
					vcpu->cpu = -1;
				}
				mutex_unlock(&vcpu->mutex);
			}
		}
	spin_unlock(&kvm_lock);
}

806 807 808 809 810 811 812 813 814 815
int kvm_dev_ioctl_check_extension(long ext)
{
	int r;

	switch (ext) {
	case KVM_CAP_IRQCHIP:
	case KVM_CAP_HLT:
	case KVM_CAP_MMU_SHADOW_CACHE_CONTROL:
	case KVM_CAP_USER_MEMORY:
	case KVM_CAP_SET_TSS_ADDR:
816
	case KVM_CAP_EXT_CPUID:
817
	case KVM_CAP_CLOCKSOURCE:
S
Sheng Yang 已提交
818
	case KVM_CAP_PIT:
819
	case KVM_CAP_NOP_IO_DELAY:
820
	case KVM_CAP_MP_STATE:
821 822
		r = 1;
		break;
823 824 825
	case KVM_CAP_VAPIC:
		r = !kvm_x86_ops->cpu_has_accelerated_tpr();
		break;
826 827 828
	case KVM_CAP_NR_VCPUS:
		r = KVM_MAX_VCPUS;
		break;
829 830 831
	case KVM_CAP_NR_MEMSLOTS:
		r = KVM_MEMORY_SLOTS;
		break;
832 833 834
	case KVM_CAP_PV_MMU:
		r = !tdp_enabled;
		break;
835 836 837 838 839 840 841 842
	default:
		r = 0;
		break;
	}
	return r;

}

843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876
long kvm_arch_dev_ioctl(struct file *filp,
			unsigned int ioctl, unsigned long arg)
{
	void __user *argp = (void __user *)arg;
	long r;

	switch (ioctl) {
	case KVM_GET_MSR_INDEX_LIST: {
		struct kvm_msr_list __user *user_msr_list = argp;
		struct kvm_msr_list msr_list;
		unsigned n;

		r = -EFAULT;
		if (copy_from_user(&msr_list, user_msr_list, sizeof msr_list))
			goto out;
		n = msr_list.nmsrs;
		msr_list.nmsrs = num_msrs_to_save + ARRAY_SIZE(emulated_msrs);
		if (copy_to_user(user_msr_list, &msr_list, sizeof msr_list))
			goto out;
		r = -E2BIG;
		if (n < num_msrs_to_save)
			goto out;
		r = -EFAULT;
		if (copy_to_user(user_msr_list->indices, &msrs_to_save,
				 num_msrs_to_save * sizeof(u32)))
			goto out;
		if (copy_to_user(user_msr_list->indices
				 + num_msrs_to_save * sizeof(u32),
				 &emulated_msrs,
				 ARRAY_SIZE(emulated_msrs) * sizeof(u32)))
			goto out;
		r = 0;
		break;
	}
877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894
	case KVM_GET_SUPPORTED_CPUID: {
		struct kvm_cpuid2 __user *cpuid_arg = argp;
		struct kvm_cpuid2 cpuid;

		r = -EFAULT;
		if (copy_from_user(&cpuid, cpuid_arg, sizeof cpuid))
			goto out;
		r = kvm_dev_ioctl_get_supported_cpuid(&cpuid,
			cpuid_arg->entries);
		if (r)
			goto out;

		r = -EFAULT;
		if (copy_to_user(cpuid_arg, &cpuid, sizeof cpuid))
			goto out;
		r = 0;
		break;
	}
895 896 897 898 899 900 901
	default:
		r = -EINVAL;
	}
out:
	return r;
}

902 903 904
void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
{
	kvm_x86_ops->vcpu_load(vcpu, cpu);
905
	kvm_write_guest_time(vcpu);
906 907 908 909 910
}

void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
{
	kvm_x86_ops->vcpu_put(vcpu);
911
	kvm_put_guest_fpu(vcpu);
912 913
}

914
static int is_efer_nx(void)
915 916 917 918
{
	u64 efer;

	rdmsrl(MSR_EFER, efer);
919 920 921 922 923 924 925 926
	return efer & EFER_NX;
}

static void cpuid_fix_nx_cap(struct kvm_vcpu *vcpu)
{
	int i;
	struct kvm_cpuid_entry2 *e, *entry;

927
	entry = NULL;
928 929
	for (i = 0; i < vcpu->arch.cpuid_nent; ++i) {
		e = &vcpu->arch.cpuid_entries[i];
930 931 932 933 934
		if (e->function == 0x80000001) {
			entry = e;
			break;
		}
	}
935
	if (entry && (entry->edx & (1 << 20)) && !is_efer_nx()) {
936 937 938 939 940
		entry->edx &= ~(1 << 20);
		printk(KERN_INFO "kvm: guest NX capability removed\n");
	}
}

941
/* when an old userspace process fills a new kernel module */
942 943 944
static int kvm_vcpu_ioctl_set_cpuid(struct kvm_vcpu *vcpu,
				    struct kvm_cpuid *cpuid,
				    struct kvm_cpuid_entry __user *entries)
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960
{
	int r, i;
	struct kvm_cpuid_entry *cpuid_entries;

	r = -E2BIG;
	if (cpuid->nent > KVM_MAX_CPUID_ENTRIES)
		goto out;
	r = -ENOMEM;
	cpuid_entries = vmalloc(sizeof(struct kvm_cpuid_entry) * cpuid->nent);
	if (!cpuid_entries)
		goto out;
	r = -EFAULT;
	if (copy_from_user(cpuid_entries, entries,
			   cpuid->nent * sizeof(struct kvm_cpuid_entry)))
		goto out_free;
	for (i = 0; i < cpuid->nent; i++) {
961 962 963 964 965 966 967 968 969 970 971 972
		vcpu->arch.cpuid_entries[i].function = cpuid_entries[i].function;
		vcpu->arch.cpuid_entries[i].eax = cpuid_entries[i].eax;
		vcpu->arch.cpuid_entries[i].ebx = cpuid_entries[i].ebx;
		vcpu->arch.cpuid_entries[i].ecx = cpuid_entries[i].ecx;
		vcpu->arch.cpuid_entries[i].edx = cpuid_entries[i].edx;
		vcpu->arch.cpuid_entries[i].index = 0;
		vcpu->arch.cpuid_entries[i].flags = 0;
		vcpu->arch.cpuid_entries[i].padding[0] = 0;
		vcpu->arch.cpuid_entries[i].padding[1] = 0;
		vcpu->arch.cpuid_entries[i].padding[2] = 0;
	}
	vcpu->arch.cpuid_nent = cpuid->nent;
973 974 975 976 977 978 979 980 981 982 983 984
	cpuid_fix_nx_cap(vcpu);
	r = 0;

out_free:
	vfree(cpuid_entries);
out:
	return r;
}

static int kvm_vcpu_ioctl_set_cpuid2(struct kvm_vcpu *vcpu,
				    struct kvm_cpuid2 *cpuid,
				    struct kvm_cpuid_entry2 __user *entries)
985 986 987 988 989 990 991
{
	int r;

	r = -E2BIG;
	if (cpuid->nent > KVM_MAX_CPUID_ENTRIES)
		goto out;
	r = -EFAULT;
992
	if (copy_from_user(&vcpu->arch.cpuid_entries, entries,
993
			   cpuid->nent * sizeof(struct kvm_cpuid_entry2)))
994
		goto out;
995
	vcpu->arch.cpuid_nent = cpuid->nent;
996 997 998 999 1000 1001
	return 0;

out:
	return r;
}

1002 1003 1004 1005 1006 1007 1008
static int kvm_vcpu_ioctl_get_cpuid2(struct kvm_vcpu *vcpu,
				    struct kvm_cpuid2 *cpuid,
				    struct kvm_cpuid_entry2 __user *entries)
{
	int r;

	r = -E2BIG;
1009
	if (cpuid->nent < vcpu->arch.cpuid_nent)
1010 1011
		goto out;
	r = -EFAULT;
1012 1013
	if (copy_to_user(entries, &vcpu->arch.cpuid_entries,
			   vcpu->arch.cpuid_nent * sizeof(struct kvm_cpuid_entry2)))
1014 1015 1016 1017
		goto out;
	return 0;

out:
1018
	cpuid->nent = vcpu->arch.cpuid_nent;
1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
	return r;
}

static inline u32 bit(int bitno)
{
	return 1 << (bitno & 31);
}

static void do_cpuid_1_ent(struct kvm_cpuid_entry2 *entry, u32 function,
			  u32 index)
{
	entry->function = function;
	entry->index = index;
	cpuid_count(entry->function, entry->index,
		&entry->eax, &entry->ebx, &entry->ecx, &entry->edx);
	entry->flags = 0;
}

static void do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
			 u32 index, int *nent, int maxnent)
{
	const u32 kvm_supported_word0_x86_features = bit(X86_FEATURE_FPU) |
		bit(X86_FEATURE_VME) | bit(X86_FEATURE_DE) |
		bit(X86_FEATURE_PSE) | bit(X86_FEATURE_TSC) |
		bit(X86_FEATURE_MSR) | bit(X86_FEATURE_PAE) |
		bit(X86_FEATURE_CX8) | bit(X86_FEATURE_APIC) |
		bit(X86_FEATURE_SEP) | bit(X86_FEATURE_PGE) |
		bit(X86_FEATURE_CMOV) | bit(X86_FEATURE_PSE36) |
		bit(X86_FEATURE_CLFLSH) | bit(X86_FEATURE_MMX) |
		bit(X86_FEATURE_FXSR) | bit(X86_FEATURE_XMM) |
		bit(X86_FEATURE_XMM2) | bit(X86_FEATURE_SELFSNOOP);
	const u32 kvm_supported_word1_x86_features = bit(X86_FEATURE_FPU) |
		bit(X86_FEATURE_VME) | bit(X86_FEATURE_DE) |
		bit(X86_FEATURE_PSE) | bit(X86_FEATURE_TSC) |
		bit(X86_FEATURE_MSR) | bit(X86_FEATURE_PAE) |
		bit(X86_FEATURE_CX8) | bit(X86_FEATURE_APIC) |
		bit(X86_FEATURE_PGE) |
		bit(X86_FEATURE_CMOV) | bit(X86_FEATURE_PSE36) |
		bit(X86_FEATURE_MMX) | bit(X86_FEATURE_FXSR) |
		bit(X86_FEATURE_SYSCALL) |
		(bit(X86_FEATURE_NX) && is_efer_nx()) |
#ifdef CONFIG_X86_64
		bit(X86_FEATURE_LM) |
#endif
		bit(X86_FEATURE_MMXEXT) |
		bit(X86_FEATURE_3DNOWEXT) |
		bit(X86_FEATURE_3DNOW);
	const u32 kvm_supported_word3_x86_features =
		bit(X86_FEATURE_XMM3) | bit(X86_FEATURE_CX16);
	const u32 kvm_supported_word6_x86_features =
		bit(X86_FEATURE_LAHF_LM) | bit(X86_FEATURE_CMP_LEGACY);

	/* all func 2 cpuid_count() should be called on the same cpu */
	get_cpu();
	do_cpuid_1_ent(entry, function, index);
	++*nent;

	switch (function) {
	case 0:
		entry->eax = min(entry->eax, (u32)0xb);
		break;
	case 1:
		entry->edx &= kvm_supported_word0_x86_features;
		entry->ecx &= kvm_supported_word3_x86_features;
		break;
	/* function 2 entries are STATEFUL. That is, repeated cpuid commands
	 * may return different values. This forces us to get_cpu() before
	 * issuing the first command, and also to emulate this annoying behavior
	 * in kvm_emulate_cpuid() using KVM_CPUID_FLAG_STATE_READ_NEXT */
	case 2: {
		int t, times = entry->eax & 0xff;

		entry->flags |= KVM_CPUID_FLAG_STATEFUL_FUNC;
		for (t = 1; t < times && *nent < maxnent; ++t) {
			do_cpuid_1_ent(&entry[t], function, 0);
			entry[t].flags |= KVM_CPUID_FLAG_STATEFUL_FUNC;
			++*nent;
		}
		break;
	}
	/* function 4 and 0xb have additional index. */
	case 4: {
1101
		int i, cache_type;
1102 1103 1104

		entry->flags |= KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
		/* read more entries until cache_type is zero */
1105 1106
		for (i = 1; *nent < maxnent; ++i) {
			cache_type = entry[i - 1].eax & 0x1f;
1107 1108
			if (!cache_type)
				break;
1109 1110
			do_cpuid_1_ent(&entry[i], function, i);
			entry[i].flags |=
1111 1112 1113 1114 1115 1116
			       KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
			++*nent;
		}
		break;
	}
	case 0xb: {
1117
		int i, level_type;
1118 1119 1120

		entry->flags |= KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
		/* read more entries until level_type is zero */
1121 1122
		for (i = 1; *nent < maxnent; ++i) {
			level_type = entry[i - 1].ecx & 0xff;
1123 1124
			if (!level_type)
				break;
1125 1126
			do_cpuid_1_ent(&entry[i], function, i);
			entry[i].flags |=
1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142
			       KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
			++*nent;
		}
		break;
	}
	case 0x80000000:
		entry->eax = min(entry->eax, 0x8000001a);
		break;
	case 0x80000001:
		entry->edx &= kvm_supported_word1_x86_features;
		entry->ecx &= kvm_supported_word6_x86_features;
		break;
	}
	put_cpu();
}

1143
static int kvm_dev_ioctl_get_supported_cpuid(struct kvm_cpuid2 *cpuid,
1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183
				    struct kvm_cpuid_entry2 __user *entries)
{
	struct kvm_cpuid_entry2 *cpuid_entries;
	int limit, nent = 0, r = -E2BIG;
	u32 func;

	if (cpuid->nent < 1)
		goto out;
	r = -ENOMEM;
	cpuid_entries = vmalloc(sizeof(struct kvm_cpuid_entry2) * cpuid->nent);
	if (!cpuid_entries)
		goto out;

	do_cpuid_ent(&cpuid_entries[0], 0, 0, &nent, cpuid->nent);
	limit = cpuid_entries[0].eax;
	for (func = 1; func <= limit && nent < cpuid->nent; ++func)
		do_cpuid_ent(&cpuid_entries[nent], func, 0,
				&nent, cpuid->nent);
	r = -E2BIG;
	if (nent >= cpuid->nent)
		goto out_free;

	do_cpuid_ent(&cpuid_entries[nent], 0x80000000, 0, &nent, cpuid->nent);
	limit = cpuid_entries[nent - 1].eax;
	for (func = 0x80000001; func <= limit && nent < cpuid->nent; ++func)
		do_cpuid_ent(&cpuid_entries[nent], func, 0,
			       &nent, cpuid->nent);
	r = -EFAULT;
	if (copy_to_user(entries, cpuid_entries,
			nent * sizeof(struct kvm_cpuid_entry2)))
		goto out_free;
	cpuid->nent = nent;
	r = 0;

out_free:
	vfree(cpuid_entries);
out:
	return r;
}

1184 1185 1186 1187
static int kvm_vcpu_ioctl_get_lapic(struct kvm_vcpu *vcpu,
				    struct kvm_lapic_state *s)
{
	vcpu_load(vcpu);
1188
	memcpy(s->regs, vcpu->arch.apic->regs, sizeof *s);
1189 1190 1191 1192 1193 1194 1195 1196 1197
	vcpu_put(vcpu);

	return 0;
}

static int kvm_vcpu_ioctl_set_lapic(struct kvm_vcpu *vcpu,
				    struct kvm_lapic_state *s)
{
	vcpu_load(vcpu);
1198
	memcpy(vcpu->arch.apic->regs, s->regs, sizeof *s);
1199 1200 1201 1202 1203 1204
	kvm_apic_post_state_restore(vcpu);
	vcpu_put(vcpu);

	return 0;
}

1205 1206 1207 1208 1209 1210 1211 1212 1213
static int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu,
				    struct kvm_interrupt *irq)
{
	if (irq->irq < 0 || irq->irq >= 256)
		return -EINVAL;
	if (irqchip_in_kernel(vcpu->kvm))
		return -ENXIO;
	vcpu_load(vcpu);

1214 1215
	set_bit(irq->irq, vcpu->arch.irq_pending);
	set_bit(irq->irq / BITS_PER_LONG, &vcpu->arch.irq_summary);
1216 1217 1218 1219 1220 1221

	vcpu_put(vcpu);

	return 0;
}

1222 1223 1224 1225 1226 1227 1228 1229 1230
static int vcpu_ioctl_tpr_access_reporting(struct kvm_vcpu *vcpu,
					   struct kvm_tpr_access_ctl *tac)
{
	if (tac->flags)
		return -EINVAL;
	vcpu->arch.tpr_access_reporting = !!tac->enabled;
	return 0;
}

1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263
long kvm_arch_vcpu_ioctl(struct file *filp,
			 unsigned int ioctl, unsigned long arg)
{
	struct kvm_vcpu *vcpu = filp->private_data;
	void __user *argp = (void __user *)arg;
	int r;

	switch (ioctl) {
	case KVM_GET_LAPIC: {
		struct kvm_lapic_state lapic;

		memset(&lapic, 0, sizeof lapic);
		r = kvm_vcpu_ioctl_get_lapic(vcpu, &lapic);
		if (r)
			goto out;
		r = -EFAULT;
		if (copy_to_user(argp, &lapic, sizeof lapic))
			goto out;
		r = 0;
		break;
	}
	case KVM_SET_LAPIC: {
		struct kvm_lapic_state lapic;

		r = -EFAULT;
		if (copy_from_user(&lapic, argp, sizeof lapic))
			goto out;
		r = kvm_vcpu_ioctl_set_lapic(vcpu, &lapic);;
		if (r)
			goto out;
		r = 0;
		break;
	}
1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275
	case KVM_INTERRUPT: {
		struct kvm_interrupt irq;

		r = -EFAULT;
		if (copy_from_user(&irq, argp, sizeof irq))
			goto out;
		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
		if (r)
			goto out;
		r = 0;
		break;
	}
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287
	case KVM_SET_CPUID: {
		struct kvm_cpuid __user *cpuid_arg = argp;
		struct kvm_cpuid cpuid;

		r = -EFAULT;
		if (copy_from_user(&cpuid, cpuid_arg, sizeof cpuid))
			goto out;
		r = kvm_vcpu_ioctl_set_cpuid(vcpu, &cpuid, cpuid_arg->entries);
		if (r)
			goto out;
		break;
	}
1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317
	case KVM_SET_CPUID2: {
		struct kvm_cpuid2 __user *cpuid_arg = argp;
		struct kvm_cpuid2 cpuid;

		r = -EFAULT;
		if (copy_from_user(&cpuid, cpuid_arg, sizeof cpuid))
			goto out;
		r = kvm_vcpu_ioctl_set_cpuid2(vcpu, &cpuid,
				cpuid_arg->entries);
		if (r)
			goto out;
		break;
	}
	case KVM_GET_CPUID2: {
		struct kvm_cpuid2 __user *cpuid_arg = argp;
		struct kvm_cpuid2 cpuid;

		r = -EFAULT;
		if (copy_from_user(&cpuid, cpuid_arg, sizeof cpuid))
			goto out;
		r = kvm_vcpu_ioctl_get_cpuid2(vcpu, &cpuid,
				cpuid_arg->entries);
		if (r)
			goto out;
		r = -EFAULT;
		if (copy_to_user(cpuid_arg, &cpuid, sizeof cpuid))
			goto out;
		r = 0;
		break;
	}
1318 1319 1320 1321 1322 1323
	case KVM_GET_MSRS:
		r = msr_io(vcpu, argp, kvm_get_msr, 1);
		break;
	case KVM_SET_MSRS:
		r = msr_io(vcpu, argp, do_set_msr, 0);
		break;
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
	case KVM_TPR_ACCESS_REPORTING: {
		struct kvm_tpr_access_ctl tac;

		r = -EFAULT;
		if (copy_from_user(&tac, argp, sizeof tac))
			goto out;
		r = vcpu_ioctl_tpr_access_reporting(vcpu, &tac);
		if (r)
			goto out;
		r = -EFAULT;
		if (copy_to_user(argp, &tac, sizeof tac))
			goto out;
		r = 0;
		break;
	};
A
Avi Kivity 已提交
1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
	case KVM_SET_VAPIC_ADDR: {
		struct kvm_vapic_addr va;

		r = -EINVAL;
		if (!irqchip_in_kernel(vcpu->kvm))
			goto out;
		r = -EFAULT;
		if (copy_from_user(&va, argp, sizeof va))
			goto out;
		r = 0;
		kvm_lapic_set_vapic_addr(vcpu, va.vapic_addr);
		break;
	}
1352 1353 1354 1355 1356 1357 1358
	default:
		r = -EINVAL;
	}
out:
	return r;
}

1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374
static int kvm_vm_ioctl_set_tss_addr(struct kvm *kvm, unsigned long addr)
{
	int ret;

	if (addr > (unsigned int)(-3 * PAGE_SIZE))
		return -1;
	ret = kvm_x86_ops->set_tss_addr(kvm, addr);
	return ret;
}

static int kvm_vm_ioctl_set_nr_mmu_pages(struct kvm *kvm,
					  u32 kvm_nr_mmu_pages)
{
	if (kvm_nr_mmu_pages < KVM_MIN_ALLOC_MMU_PAGES)
		return -EINVAL;

1375
	down_write(&kvm->slots_lock);
1376 1377

	kvm_mmu_change_mmu_pages(kvm, kvm_nr_mmu_pages);
1378
	kvm->arch.n_requested_mmu_pages = kvm_nr_mmu_pages;
1379

1380
	up_write(&kvm->slots_lock);
1381 1382 1383 1384 1385
	return 0;
}

static int kvm_vm_ioctl_get_nr_mmu_pages(struct kvm *kvm)
{
1386
	return kvm->arch.n_alloc_mmu_pages;
1387 1388
}

1389 1390 1391 1392 1393
gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
{
	int i;
	struct kvm_mem_alias *alias;

1394 1395
	for (i = 0; i < kvm->arch.naliases; ++i) {
		alias = &kvm->arch.aliases[i];
1396 1397 1398 1399 1400 1401 1402
		if (gfn >= alias->base_gfn
		    && gfn < alias->base_gfn + alias->npages)
			return alias->target_gfn + gfn - alias->base_gfn;
	}
	return gfn;
}

1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428
/*
 * Set a new alias region.  Aliases map a portion of physical memory into
 * another portion.  This is useful for memory windows, for example the PC
 * VGA region.
 */
static int kvm_vm_ioctl_set_memory_alias(struct kvm *kvm,
					 struct kvm_memory_alias *alias)
{
	int r, n;
	struct kvm_mem_alias *p;

	r = -EINVAL;
	/* General sanity checks */
	if (alias->memory_size & (PAGE_SIZE - 1))
		goto out;
	if (alias->guest_phys_addr & (PAGE_SIZE - 1))
		goto out;
	if (alias->slot >= KVM_ALIAS_SLOTS)
		goto out;
	if (alias->guest_phys_addr + alias->memory_size
	    < alias->guest_phys_addr)
		goto out;
	if (alias->target_phys_addr + alias->memory_size
	    < alias->target_phys_addr)
		goto out;

1429
	down_write(&kvm->slots_lock);
1430

1431
	p = &kvm->arch.aliases[alias->slot];
1432 1433 1434 1435 1436
	p->base_gfn = alias->guest_phys_addr >> PAGE_SHIFT;
	p->npages = alias->memory_size >> PAGE_SHIFT;
	p->target_gfn = alias->target_phys_addr >> PAGE_SHIFT;

	for (n = KVM_ALIAS_SLOTS; n > 0; --n)
1437
		if (kvm->arch.aliases[n - 1].npages)
1438
			break;
1439
	kvm->arch.naliases = n;
1440 1441 1442

	kvm_mmu_zap_all(kvm);

1443
	up_write(&kvm->slots_lock);
1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507

	return 0;

out:
	return r;
}

static int kvm_vm_ioctl_get_irqchip(struct kvm *kvm, struct kvm_irqchip *chip)
{
	int r;

	r = 0;
	switch (chip->chip_id) {
	case KVM_IRQCHIP_PIC_MASTER:
		memcpy(&chip->chip.pic,
			&pic_irqchip(kvm)->pics[0],
			sizeof(struct kvm_pic_state));
		break;
	case KVM_IRQCHIP_PIC_SLAVE:
		memcpy(&chip->chip.pic,
			&pic_irqchip(kvm)->pics[1],
			sizeof(struct kvm_pic_state));
		break;
	case KVM_IRQCHIP_IOAPIC:
		memcpy(&chip->chip.ioapic,
			ioapic_irqchip(kvm),
			sizeof(struct kvm_ioapic_state));
		break;
	default:
		r = -EINVAL;
		break;
	}
	return r;
}

static int kvm_vm_ioctl_set_irqchip(struct kvm *kvm, struct kvm_irqchip *chip)
{
	int r;

	r = 0;
	switch (chip->chip_id) {
	case KVM_IRQCHIP_PIC_MASTER:
		memcpy(&pic_irqchip(kvm)->pics[0],
			&chip->chip.pic,
			sizeof(struct kvm_pic_state));
		break;
	case KVM_IRQCHIP_PIC_SLAVE:
		memcpy(&pic_irqchip(kvm)->pics[1],
			&chip->chip.pic,
			sizeof(struct kvm_pic_state));
		break;
	case KVM_IRQCHIP_IOAPIC:
		memcpy(ioapic_irqchip(kvm),
			&chip->chip.ioapic,
			sizeof(struct kvm_ioapic_state));
		break;
	default:
		r = -EINVAL;
		break;
	}
	kvm_pic_update_irq(pic_irqchip(kvm));
	return r;
}

1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524
static int kvm_vm_ioctl_get_pit(struct kvm *kvm, struct kvm_pit_state *ps)
{
	int r = 0;

	memcpy(ps, &kvm->arch.vpit->pit_state, sizeof(struct kvm_pit_state));
	return r;
}

static int kvm_vm_ioctl_set_pit(struct kvm *kvm, struct kvm_pit_state *ps)
{
	int r = 0;

	memcpy(&kvm->arch.vpit->pit_state, ps, sizeof(struct kvm_pit_state));
	kvm_pit_load_count(kvm, 0, ps->channels[0].count);
	return r;
}

1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535
/*
 * Get (and clear) the dirty memory log for a memory slot.
 */
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
				      struct kvm_dirty_log *log)
{
	int r;
	int n;
	struct kvm_memory_slot *memslot;
	int is_dirty = 0;

1536
	down_write(&kvm->slots_lock);
1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551

	r = kvm_get_dirty_log(kvm, log, &is_dirty);
	if (r)
		goto out;

	/* If nothing is dirty, don't bother messing with page tables. */
	if (is_dirty) {
		kvm_mmu_slot_remove_write_access(kvm, log->slot);
		kvm_flush_remote_tlbs(kvm);
		memslot = &kvm->memslots[log->slot];
		n = ALIGN(memslot->npages, BITS_PER_LONG) / 8;
		memset(memslot->dirty_bitmap, 0, n);
	}
	r = 0;
out:
1552
	up_write(&kvm->slots_lock);
1553 1554 1555
	return r;
}

1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605
long kvm_arch_vm_ioctl(struct file *filp,
		       unsigned int ioctl, unsigned long arg)
{
	struct kvm *kvm = filp->private_data;
	void __user *argp = (void __user *)arg;
	int r = -EINVAL;

	switch (ioctl) {
	case KVM_SET_TSS_ADDR:
		r = kvm_vm_ioctl_set_tss_addr(kvm, arg);
		if (r < 0)
			goto out;
		break;
	case KVM_SET_MEMORY_REGION: {
		struct kvm_memory_region kvm_mem;
		struct kvm_userspace_memory_region kvm_userspace_mem;

		r = -EFAULT;
		if (copy_from_user(&kvm_mem, argp, sizeof kvm_mem))
			goto out;
		kvm_userspace_mem.slot = kvm_mem.slot;
		kvm_userspace_mem.flags = kvm_mem.flags;
		kvm_userspace_mem.guest_phys_addr = kvm_mem.guest_phys_addr;
		kvm_userspace_mem.memory_size = kvm_mem.memory_size;
		r = kvm_vm_ioctl_set_memory_region(kvm, &kvm_userspace_mem, 0);
		if (r)
			goto out;
		break;
	}
	case KVM_SET_NR_MMU_PAGES:
		r = kvm_vm_ioctl_set_nr_mmu_pages(kvm, arg);
		if (r)
			goto out;
		break;
	case KVM_GET_NR_MMU_PAGES:
		r = kvm_vm_ioctl_get_nr_mmu_pages(kvm);
		break;
	case KVM_SET_MEMORY_ALIAS: {
		struct kvm_memory_alias alias;

		r = -EFAULT;
		if (copy_from_user(&alias, argp, sizeof alias))
			goto out;
		r = kvm_vm_ioctl_set_memory_alias(kvm, &alias);
		if (r)
			goto out;
		break;
	}
	case KVM_CREATE_IRQCHIP:
		r = -ENOMEM;
1606 1607
		kvm->arch.vpic = kvm_create_pic(kvm);
		if (kvm->arch.vpic) {
1608 1609
			r = kvm_ioapic_init(kvm);
			if (r) {
1610 1611
				kfree(kvm->arch.vpic);
				kvm->arch.vpic = NULL;
1612 1613 1614 1615 1616
				goto out;
			}
		} else
			goto out;
		break;
S
Sheng Yang 已提交
1617 1618 1619 1620 1621 1622
	case KVM_CREATE_PIT:
		r = -ENOMEM;
		kvm->arch.vpit = kvm_create_pit(kvm);
		if (kvm->arch.vpit)
			r = 0;
		break;
1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634
	case KVM_IRQ_LINE: {
		struct kvm_irq_level irq_event;

		r = -EFAULT;
		if (copy_from_user(&irq_event, argp, sizeof irq_event))
			goto out;
		if (irqchip_in_kernel(kvm)) {
			mutex_lock(&kvm->lock);
			if (irq_event.irq < 16)
				kvm_pic_set_irq(pic_irqchip(kvm),
					irq_event.irq,
					irq_event.level);
1635
			kvm_ioapic_set_irq(kvm->arch.vioapic,
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
					irq_event.irq,
					irq_event.level);
			mutex_unlock(&kvm->lock);
			r = 0;
		}
		break;
	}
	case KVM_GET_IRQCHIP: {
		/* 0: PIC master, 1: PIC slave, 2: IOAPIC */
		struct kvm_irqchip chip;

		r = -EFAULT;
		if (copy_from_user(&chip, argp, sizeof chip))
			goto out;
		r = -ENXIO;
		if (!irqchip_in_kernel(kvm))
			goto out;
		r = kvm_vm_ioctl_get_irqchip(kvm, &chip);
		if (r)
			goto out;
		r = -EFAULT;
		if (copy_to_user(argp, &chip, sizeof chip))
			goto out;
		r = 0;
		break;
	}
	case KVM_SET_IRQCHIP: {
		/* 0: PIC master, 1: PIC slave, 2: IOAPIC */
		struct kvm_irqchip chip;

		r = -EFAULT;
		if (copy_from_user(&chip, argp, sizeof chip))
			goto out;
		r = -ENXIO;
		if (!irqchip_in_kernel(kvm))
			goto out;
		r = kvm_vm_ioctl_set_irqchip(kvm, &chip);
		if (r)
			goto out;
		r = 0;
		break;
	}
1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708
	case KVM_GET_PIT: {
		struct kvm_pit_state ps;
		r = -EFAULT;
		if (copy_from_user(&ps, argp, sizeof ps))
			goto out;
		r = -ENXIO;
		if (!kvm->arch.vpit)
			goto out;
		r = kvm_vm_ioctl_get_pit(kvm, &ps);
		if (r)
			goto out;
		r = -EFAULT;
		if (copy_to_user(argp, &ps, sizeof ps))
			goto out;
		r = 0;
		break;
	}
	case KVM_SET_PIT: {
		struct kvm_pit_state ps;
		r = -EFAULT;
		if (copy_from_user(&ps, argp, sizeof ps))
			goto out;
		r = -ENXIO;
		if (!kvm->arch.vpit)
			goto out;
		r = kvm_vm_ioctl_set_pit(kvm, &ps);
		if (r)
			goto out;
		r = 0;
		break;
	}
1709 1710 1711 1712 1713 1714 1715
	default:
		;
	}
out:
	return r;
}

1716
static void kvm_init_msr_list(void)
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730
{
	u32 dummy[2];
	unsigned i, j;

	for (i = j = 0; i < ARRAY_SIZE(msrs_to_save); i++) {
		if (rdmsr_safe(msrs_to_save[i], &dummy[0], &dummy[1]) < 0)
			continue;
		if (j < i)
			msrs_to_save[j] = msrs_to_save[i];
		j++;
	}
	num_msrs_to_save = j;
}

1731 1732 1733 1734 1735 1736 1737 1738
/*
 * Only apic need an MMIO device hook, so shortcut now..
 */
static struct kvm_io_device *vcpu_find_pervcpu_dev(struct kvm_vcpu *vcpu,
						gpa_t addr)
{
	struct kvm_io_device *dev;

1739 1740
	if (vcpu->arch.apic) {
		dev = &vcpu->arch.apic->dev;
1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764
		if (dev->in_range(dev, addr))
			return dev;
	}
	return NULL;
}


static struct kvm_io_device *vcpu_find_mmio_dev(struct kvm_vcpu *vcpu,
						gpa_t addr)
{
	struct kvm_io_device *dev;

	dev = vcpu_find_pervcpu_dev(vcpu, addr);
	if (dev == NULL)
		dev = kvm_io_bus_find_dev(&vcpu->kvm->mmio_bus, addr);
	return dev;
}

int emulator_read_std(unsigned long addr,
			     void *val,
			     unsigned int bytes,
			     struct kvm_vcpu *vcpu)
{
	void *data = val;
1765
	int r = X86EMUL_CONTINUE;
1766 1767

	while (bytes) {
1768
		gpa_t gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);
1769 1770 1771 1772
		unsigned offset = addr & (PAGE_SIZE-1);
		unsigned tocopy = min(bytes, (unsigned)PAGE_SIZE - offset);
		int ret;

1773 1774 1775 1776
		if (gpa == UNMAPPED_GVA) {
			r = X86EMUL_PROPAGATE_FAULT;
			goto out;
		}
1777
		ret = kvm_read_guest(vcpu->kvm, gpa, data, tocopy);
1778 1779 1780 1781
		if (ret < 0) {
			r = X86EMUL_UNHANDLEABLE;
			goto out;
		}
1782 1783 1784 1785 1786

		bytes -= tocopy;
		data += tocopy;
		addr += tocopy;
	}
1787 1788
out:
	return r;
1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805
}
EXPORT_SYMBOL_GPL(emulator_read_std);

static int emulator_read_emulated(unsigned long addr,
				  void *val,
				  unsigned int bytes,
				  struct kvm_vcpu *vcpu)
{
	struct kvm_io_device *mmio_dev;
	gpa_t                 gpa;

	if (vcpu->mmio_read_completed) {
		memcpy(val, vcpu->mmio_data, bytes);
		vcpu->mmio_read_completed = 0;
		return X86EMUL_CONTINUE;
	}

1806
	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);
1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821

	/* For APIC access vmexit */
	if ((gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE)
		goto mmio;

	if (emulator_read_std(addr, val, bytes, vcpu)
			== X86EMUL_CONTINUE)
		return X86EMUL_CONTINUE;
	if (gpa == UNMAPPED_GVA)
		return X86EMUL_PROPAGATE_FAULT;

mmio:
	/*
	 * Is this MMIO handled locally?
	 */
1822
	mutex_lock(&vcpu->kvm->lock);
1823 1824 1825
	mmio_dev = vcpu_find_mmio_dev(vcpu, gpa);
	if (mmio_dev) {
		kvm_iodevice_read(mmio_dev, gpa, bytes, val);
1826
		mutex_unlock(&vcpu->kvm->lock);
1827 1828
		return X86EMUL_CONTINUE;
	}
1829
	mutex_unlock(&vcpu->kvm->lock);
1830 1831 1832 1833 1834 1835 1836 1837 1838

	vcpu->mmio_needed = 1;
	vcpu->mmio_phys_addr = gpa;
	vcpu->mmio_size = bytes;
	vcpu->mmio_is_write = 0;

	return X86EMUL_UNHANDLEABLE;
}

1839
int emulator_write_phys(struct kvm_vcpu *vcpu, gpa_t gpa,
1840
			  const void *val, int bytes)
1841 1842 1843 1844
{
	int ret;

	ret = kvm_write_guest(vcpu->kvm, gpa, val, bytes);
1845
	if (ret < 0)
1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856
		return 0;
	kvm_mmu_pte_write(vcpu, gpa, val, bytes);
	return 1;
}

static int emulator_write_emulated_onepage(unsigned long addr,
					   const void *val,
					   unsigned int bytes,
					   struct kvm_vcpu *vcpu)
{
	struct kvm_io_device *mmio_dev;
1857 1858 1859
	gpa_t                 gpa;

	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);
1860 1861

	if (gpa == UNMAPPED_GVA) {
1862
		kvm_inject_page_fault(vcpu, addr, 2);
1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876
		return X86EMUL_PROPAGATE_FAULT;
	}

	/* For APIC access vmexit */
	if ((gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE)
		goto mmio;

	if (emulator_write_phys(vcpu, gpa, val, bytes))
		return X86EMUL_CONTINUE;

mmio:
	/*
	 * Is this MMIO handled locally?
	 */
1877
	mutex_lock(&vcpu->kvm->lock);
1878 1879 1880
	mmio_dev = vcpu_find_mmio_dev(vcpu, gpa);
	if (mmio_dev) {
		kvm_iodevice_write(mmio_dev, gpa, bytes, val);
1881
		mutex_unlock(&vcpu->kvm->lock);
1882 1883
		return X86EMUL_CONTINUE;
	}
1884
	mutex_unlock(&vcpu->kvm->lock);
1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927

	vcpu->mmio_needed = 1;
	vcpu->mmio_phys_addr = gpa;
	vcpu->mmio_size = bytes;
	vcpu->mmio_is_write = 1;
	memcpy(vcpu->mmio_data, val, bytes);

	return X86EMUL_CONTINUE;
}

int emulator_write_emulated(unsigned long addr,
				   const void *val,
				   unsigned int bytes,
				   struct kvm_vcpu *vcpu)
{
	/* Crossing a page boundary? */
	if (((addr + bytes - 1) ^ addr) & PAGE_MASK) {
		int rc, now;

		now = -addr & ~PAGE_MASK;
		rc = emulator_write_emulated_onepage(addr, val, now, vcpu);
		if (rc != X86EMUL_CONTINUE)
			return rc;
		addr += now;
		val += now;
		bytes -= now;
	}
	return emulator_write_emulated_onepage(addr, val, bytes, vcpu);
}
EXPORT_SYMBOL_GPL(emulator_write_emulated);

static int emulator_cmpxchg_emulated(unsigned long addr,
				     const void *old,
				     const void *new,
				     unsigned int bytes,
				     struct kvm_vcpu *vcpu)
{
	static int reported;

	if (!reported) {
		reported = 1;
		printk(KERN_WARNING "kvm: emulating exchange as write\n");
	}
1928 1929 1930
#ifndef CONFIG_X86_64
	/* guests cmpxchg8b have to be emulated atomically */
	if (bytes == 8) {
1931
		gpa_t gpa;
1932
		struct page *page;
A
Andrew Morton 已提交
1933
		char *kaddr;
1934 1935
		u64 val;

1936 1937
		gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, addr);

1938 1939 1940 1941 1942 1943 1944 1945
		if (gpa == UNMAPPED_GVA ||
		   (gpa & PAGE_MASK) == APIC_DEFAULT_PHYS_BASE)
			goto emul_write;

		if (((gpa + bytes - 1) & PAGE_MASK) != (gpa & PAGE_MASK))
			goto emul_write;

		val = *(u64 *)new;
1946 1947

		down_read(&current->mm->mmap_sem);
1948
		page = gfn_to_page(vcpu->kvm, gpa >> PAGE_SHIFT);
1949 1950
		up_read(&current->mm->mmap_sem);

A
Andrew Morton 已提交
1951 1952 1953
		kaddr = kmap_atomic(page, KM_USER0);
		set_64bit((u64 *)(kaddr + offset_in_page(gpa)), val);
		kunmap_atomic(kaddr, KM_USER0);
1954 1955
		kvm_release_page_dirty(page);
	}
1956
emul_write:
1957 1958
#endif

1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973
	return emulator_write_emulated(addr, new, bytes, vcpu);
}

static unsigned long get_segment_base(struct kvm_vcpu *vcpu, int seg)
{
	return kvm_x86_ops->get_segment_base(vcpu, seg);
}

int emulate_invlpg(struct kvm_vcpu *vcpu, gva_t address)
{
	return X86EMUL_CONTINUE;
}

int emulate_clts(struct kvm_vcpu *vcpu)
{
1974
	kvm_x86_ops->set_cr0(vcpu, vcpu->arch.cr0 & ~X86_CR0_TS);
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986
	return X86EMUL_CONTINUE;
}

int emulator_get_dr(struct x86_emulate_ctxt *ctxt, int dr, unsigned long *dest)
{
	struct kvm_vcpu *vcpu = ctxt->vcpu;

	switch (dr) {
	case 0 ... 3:
		*dest = kvm_x86_ops->get_dr(vcpu, dr);
		return X86EMUL_CONTINUE;
	default:
1987
		pr_unimpl(vcpu, "%s: unexpected dr %u\n", __func__, dr);
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
		return X86EMUL_UNHANDLEABLE;
	}
}

int emulator_set_dr(struct x86_emulate_ctxt *ctxt, int dr, unsigned long value)
{
	unsigned long mask = (ctxt->mode == X86EMUL_MODE_PROT64) ? ~0ULL : ~0U;
	int exception;

	kvm_x86_ops->set_dr(ctxt->vcpu, dr, value & mask, &exception);
	if (exception) {
		/* FIXME: better handling */
		return X86EMUL_UNHANDLEABLE;
	}
	return X86EMUL_CONTINUE;
}

void kvm_report_emulation_failure(struct kvm_vcpu *vcpu, const char *context)
{
	static int reported;
	u8 opcodes[4];
2009
	unsigned long rip = vcpu->arch.rip;
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024
	unsigned long rip_linear;

	rip_linear = rip + get_segment_base(vcpu, VCPU_SREG_CS);

	if (reported)
		return;

	emulator_read_std(rip_linear, (void *)opcodes, 4, vcpu);

	printk(KERN_ERR "emulation failed (%s) rip %lx %02x %02x %02x %02x\n",
	       context, rip, opcodes[0], opcodes[1], opcodes[2], opcodes[3]);
	reported = 1;
}
EXPORT_SYMBOL_GPL(kvm_report_emulation_failure);

2025
static struct x86_emulate_ops emulate_ops = {
2026 2027 2028 2029 2030 2031 2032 2033 2034 2035
	.read_std            = emulator_read_std,
	.read_emulated       = emulator_read_emulated,
	.write_emulated      = emulator_write_emulated,
	.cmpxchg_emulated    = emulator_cmpxchg_emulated,
};

int emulate_instruction(struct kvm_vcpu *vcpu,
			struct kvm_run *run,
			unsigned long cr2,
			u16 error_code,
2036
			int emulation_type)
2037 2038
{
	int r;
2039
	struct decode_cache *c;
2040

2041
	vcpu->arch.mmio_fault_cr2 = cr2;
2042 2043 2044
	kvm_x86_ops->cache_regs(vcpu);

	vcpu->mmio_is_write = 0;
2045
	vcpu->arch.pio.string = 0;
2046

2047
	if (!(emulation_type & EMULTYPE_NO_DECODE)) {
2048 2049 2050
		int cs_db, cs_l;
		kvm_x86_ops->get_cs_db_l_bits(vcpu, &cs_db, &cs_l);

2051 2052 2053 2054
		vcpu->arch.emulate_ctxt.vcpu = vcpu;
		vcpu->arch.emulate_ctxt.eflags = kvm_x86_ops->get_rflags(vcpu);
		vcpu->arch.emulate_ctxt.mode =
			(vcpu->arch.emulate_ctxt.eflags & X86_EFLAGS_VM)
2055 2056 2057 2058
			? X86EMUL_MODE_REAL : cs_l
			? X86EMUL_MODE_PROT64 :	cs_db
			? X86EMUL_MODE_PROT32 : X86EMUL_MODE_PROT16;

2059 2060 2061 2062 2063
		if (vcpu->arch.emulate_ctxt.mode == X86EMUL_MODE_PROT64) {
			vcpu->arch.emulate_ctxt.cs_base = 0;
			vcpu->arch.emulate_ctxt.ds_base = 0;
			vcpu->arch.emulate_ctxt.es_base = 0;
			vcpu->arch.emulate_ctxt.ss_base = 0;
2064
		} else {
2065
			vcpu->arch.emulate_ctxt.cs_base =
2066
					get_segment_base(vcpu, VCPU_SREG_CS);
2067
			vcpu->arch.emulate_ctxt.ds_base =
2068
					get_segment_base(vcpu, VCPU_SREG_DS);
2069
			vcpu->arch.emulate_ctxt.es_base =
2070
					get_segment_base(vcpu, VCPU_SREG_ES);
2071
			vcpu->arch.emulate_ctxt.ss_base =
2072 2073 2074
					get_segment_base(vcpu, VCPU_SREG_SS);
		}

2075
		vcpu->arch.emulate_ctxt.gs_base =
2076
					get_segment_base(vcpu, VCPU_SREG_GS);
2077
		vcpu->arch.emulate_ctxt.fs_base =
2078 2079
					get_segment_base(vcpu, VCPU_SREG_FS);

2080
		r = x86_decode_insn(&vcpu->arch.emulate_ctxt, &emulate_ops);
2081 2082 2083 2084 2085 2086 2087 2088 2089 2090

		/* Reject the instructions other than VMCALL/VMMCALL when
		 * try to emulate invalid opcode */
		c = &vcpu->arch.emulate_ctxt.decode;
		if ((emulation_type & EMULTYPE_TRAP_UD) &&
		    (!(c->twobyte && c->b == 0x01 &&
		      (c->modrm_reg == 0 || c->modrm_reg == 3) &&
		       c->modrm_mod == 3 && c->modrm_rm == 1)))
			return EMULATE_FAIL;

2091
		++vcpu->stat.insn_emulation;
2092
		if (r)  {
2093
			++vcpu->stat.insn_emulation_fail;
2094 2095 2096 2097 2098 2099
			if (kvm_mmu_unprotect_page_virt(vcpu, cr2))
				return EMULATE_DONE;
			return EMULATE_FAIL;
		}
	}

2100
	r = x86_emulate_insn(&vcpu->arch.emulate_ctxt, &emulate_ops);
2101

2102
	if (vcpu->arch.pio.string)
2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123
		return EMULATE_DO_MMIO;

	if ((r || vcpu->mmio_is_write) && run) {
		run->exit_reason = KVM_EXIT_MMIO;
		run->mmio.phys_addr = vcpu->mmio_phys_addr;
		memcpy(run->mmio.data, vcpu->mmio_data, 8);
		run->mmio.len = vcpu->mmio_size;
		run->mmio.is_write = vcpu->mmio_is_write;
	}

	if (r) {
		if (kvm_mmu_unprotect_page_virt(vcpu, cr2))
			return EMULATE_DONE;
		if (!vcpu->mmio_needed) {
			kvm_report_emulation_failure(vcpu, "mmio");
			return EMULATE_FAIL;
		}
		return EMULATE_DO_MMIO;
	}

	kvm_x86_ops->decache_regs(vcpu);
2124
	kvm_x86_ops->set_rflags(vcpu, vcpu->arch.emulate_ctxt.eflags);
2125 2126 2127 2128 2129 2130 2131 2132 2133 2134

	if (vcpu->mmio_is_write) {
		vcpu->mmio_needed = 0;
		return EMULATE_DO_MMIO;
	}

	return EMULATE_DONE;
}
EXPORT_SYMBOL_GPL(emulate_instruction);

2135 2136 2137 2138
static void free_pio_guest_pages(struct kvm_vcpu *vcpu)
{
	int i;

2139 2140 2141 2142
	for (i = 0; i < ARRAY_SIZE(vcpu->arch.pio.guest_pages); ++i)
		if (vcpu->arch.pio.guest_pages[i]) {
			kvm_release_page_dirty(vcpu->arch.pio.guest_pages[i]);
			vcpu->arch.pio.guest_pages[i] = NULL;
2143 2144 2145 2146 2147
		}
}

static int pio_copy_data(struct kvm_vcpu *vcpu)
{
2148
	void *p = vcpu->arch.pio_data;
2149 2150
	void *q;
	unsigned bytes;
2151
	int nr_pages = vcpu->arch.pio.guest_pages[1] ? 2 : 1;
2152

2153
	q = vmap(vcpu->arch.pio.guest_pages, nr_pages, VM_READ|VM_WRITE,
2154 2155 2156 2157 2158
		 PAGE_KERNEL);
	if (!q) {
		free_pio_guest_pages(vcpu);
		return -ENOMEM;
	}
2159 2160 2161
	q += vcpu->arch.pio.guest_page_offset;
	bytes = vcpu->arch.pio.size * vcpu->arch.pio.cur_count;
	if (vcpu->arch.pio.in)
2162 2163 2164
		memcpy(q, p, bytes);
	else
		memcpy(p, q, bytes);
2165
	q -= vcpu->arch.pio.guest_page_offset;
2166 2167 2168 2169 2170 2171 2172
	vunmap(q);
	free_pio_guest_pages(vcpu);
	return 0;
}

int complete_pio(struct kvm_vcpu *vcpu)
{
2173
	struct kvm_pio_request *io = &vcpu->arch.pio;
2174 2175 2176 2177 2178 2179 2180
	long delta;
	int r;

	kvm_x86_ops->cache_regs(vcpu);

	if (!io->string) {
		if (io->in)
2181
			memcpy(&vcpu->arch.regs[VCPU_REGS_RAX], vcpu->arch.pio_data,
2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198
			       io->size);
	} else {
		if (io->in) {
			r = pio_copy_data(vcpu);
			if (r) {
				kvm_x86_ops->cache_regs(vcpu);
				return r;
			}
		}

		delta = 1;
		if (io->rep) {
			delta *= io->cur_count;
			/*
			 * The size of the register should really depend on
			 * current address size.
			 */
2199
			vcpu->arch.regs[VCPU_REGS_RCX] -= delta;
2200 2201 2202 2203 2204
		}
		if (io->down)
			delta = -delta;
		delta *= io->size;
		if (io->in)
2205
			vcpu->arch.regs[VCPU_REGS_RDI] += delta;
2206
		else
2207
			vcpu->arch.regs[VCPU_REGS_RSI] += delta;
2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224
	}

	kvm_x86_ops->decache_regs(vcpu);

	io->count -= io->cur_count;
	io->cur_count = 0;

	return 0;
}

static void kernel_pio(struct kvm_io_device *pio_dev,
		       struct kvm_vcpu *vcpu,
		       void *pd)
{
	/* TODO: String I/O for in kernel device */

	mutex_lock(&vcpu->kvm->lock);
2225 2226 2227
	if (vcpu->arch.pio.in)
		kvm_iodevice_read(pio_dev, vcpu->arch.pio.port,
				  vcpu->arch.pio.size,
2228 2229
				  pd);
	else
2230 2231
		kvm_iodevice_write(pio_dev, vcpu->arch.pio.port,
				   vcpu->arch.pio.size,
2232 2233 2234 2235 2236 2237 2238
				   pd);
	mutex_unlock(&vcpu->kvm->lock);
}

static void pio_string_write(struct kvm_io_device *pio_dev,
			     struct kvm_vcpu *vcpu)
{
2239 2240
	struct kvm_pio_request *io = &vcpu->arch.pio;
	void *pd = vcpu->arch.pio_data;
2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265
	int i;

	mutex_lock(&vcpu->kvm->lock);
	for (i = 0; i < io->cur_count; i++) {
		kvm_iodevice_write(pio_dev, io->port,
				   io->size,
				   pd);
		pd += io->size;
	}
	mutex_unlock(&vcpu->kvm->lock);
}

static struct kvm_io_device *vcpu_find_pio_dev(struct kvm_vcpu *vcpu,
					       gpa_t addr)
{
	return kvm_io_bus_find_dev(&vcpu->kvm->pio_bus, addr);
}

int kvm_emulate_pio(struct kvm_vcpu *vcpu, struct kvm_run *run, int in,
		  int size, unsigned port)
{
	struct kvm_io_device *pio_dev;

	vcpu->run->exit_reason = KVM_EXIT_IO;
	vcpu->run->io.direction = in ? KVM_EXIT_IO_IN : KVM_EXIT_IO_OUT;
2266
	vcpu->run->io.size = vcpu->arch.pio.size = size;
2267
	vcpu->run->io.data_offset = KVM_PIO_PAGE_OFFSET * PAGE_SIZE;
2268 2269 2270 2271 2272 2273 2274
	vcpu->run->io.count = vcpu->arch.pio.count = vcpu->arch.pio.cur_count = 1;
	vcpu->run->io.port = vcpu->arch.pio.port = port;
	vcpu->arch.pio.in = in;
	vcpu->arch.pio.string = 0;
	vcpu->arch.pio.down = 0;
	vcpu->arch.pio.guest_page_offset = 0;
	vcpu->arch.pio.rep = 0;
2275

F
Feng (Eric) Liu 已提交
2276 2277 2278 2279 2280 2281 2282
	if (vcpu->run->io.direction == KVM_EXIT_IO_IN)
		KVMTRACE_2D(IO_READ, vcpu, vcpu->run->io.port, (u32)size,
			    handler);
	else
		KVMTRACE_2D(IO_WRITE, vcpu, vcpu->run->io.port, (u32)size,
			    handler);

2283
	kvm_x86_ops->cache_regs(vcpu);
2284
	memcpy(vcpu->arch.pio_data, &vcpu->arch.regs[VCPU_REGS_RAX], 4);
2285 2286 2287 2288 2289 2290
	kvm_x86_ops->decache_regs(vcpu);

	kvm_x86_ops->skip_emulated_instruction(vcpu);

	pio_dev = vcpu_find_pio_dev(vcpu, port);
	if (pio_dev) {
2291
		kernel_pio(pio_dev, vcpu, vcpu->arch.pio_data);
2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310
		complete_pio(vcpu);
		return 1;
	}
	return 0;
}
EXPORT_SYMBOL_GPL(kvm_emulate_pio);

int kvm_emulate_pio_string(struct kvm_vcpu *vcpu, struct kvm_run *run, int in,
		  int size, unsigned long count, int down,
		  gva_t address, int rep, unsigned port)
{
	unsigned now, in_page;
	int i, ret = 0;
	int nr_pages = 1;
	struct page *page;
	struct kvm_io_device *pio_dev;

	vcpu->run->exit_reason = KVM_EXIT_IO;
	vcpu->run->io.direction = in ? KVM_EXIT_IO_IN : KVM_EXIT_IO_OUT;
2311
	vcpu->run->io.size = vcpu->arch.pio.size = size;
2312
	vcpu->run->io.data_offset = KVM_PIO_PAGE_OFFSET * PAGE_SIZE;
2313 2314 2315 2316 2317 2318 2319
	vcpu->run->io.count = vcpu->arch.pio.count = vcpu->arch.pio.cur_count = count;
	vcpu->run->io.port = vcpu->arch.pio.port = port;
	vcpu->arch.pio.in = in;
	vcpu->arch.pio.string = 1;
	vcpu->arch.pio.down = down;
	vcpu->arch.pio.guest_page_offset = offset_in_page(address);
	vcpu->arch.pio.rep = rep;
2320

F
Feng (Eric) Liu 已提交
2321 2322 2323 2324 2325 2326 2327
	if (vcpu->run->io.direction == KVM_EXIT_IO_IN)
		KVMTRACE_2D(IO_READ, vcpu, vcpu->run->io.port, (u32)size,
			    handler);
	else
		KVMTRACE_2D(IO_WRITE, vcpu, vcpu->run->io.port, (u32)size,
			    handler);

2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351
	if (!count) {
		kvm_x86_ops->skip_emulated_instruction(vcpu);
		return 1;
	}

	if (!down)
		in_page = PAGE_SIZE - offset_in_page(address);
	else
		in_page = offset_in_page(address) + size;
	now = min(count, (unsigned long)in_page / size);
	if (!now) {
		/*
		 * String I/O straddles page boundary.  Pin two guest pages
		 * so that we satisfy atomicity constraints.  Do just one
		 * transaction to avoid complexity.
		 */
		nr_pages = 2;
		now = 1;
	}
	if (down) {
		/*
		 * String I/O in reverse.  Yuck.  Kill the guest, fix later.
		 */
		pr_unimpl(vcpu, "guest string pio down\n");
2352
		kvm_inject_gp(vcpu, 0);
2353 2354 2355
		return 1;
	}
	vcpu->run->io.count = now;
2356
	vcpu->arch.pio.cur_count = now;
2357

2358
	if (vcpu->arch.pio.cur_count == vcpu->arch.pio.count)
2359 2360 2361 2362
		kvm_x86_ops->skip_emulated_instruction(vcpu);

	for (i = 0; i < nr_pages; ++i) {
		page = gva_to_page(vcpu, address + i * PAGE_SIZE);
2363
		vcpu->arch.pio.guest_pages[i] = page;
2364
		if (!page) {
2365
			kvm_inject_gp(vcpu, 0);
2366 2367 2368 2369 2370 2371
			free_pio_guest_pages(vcpu);
			return 1;
		}
	}

	pio_dev = vcpu_find_pio_dev(vcpu, port);
2372
	if (!vcpu->arch.pio.in) {
2373 2374 2375 2376 2377
		/* string PIO write */
		ret = pio_copy_data(vcpu);
		if (ret >= 0 && pio_dev) {
			pio_string_write(pio_dev, vcpu);
			complete_pio(vcpu);
2378
			if (vcpu->arch.pio.count == 0)
2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389
				ret = 1;
		}
	} else if (pio_dev)
		pr_unimpl(vcpu, "no string pio read support yet, "
		       "port %x size %d count %ld\n",
			port, size, count);

	return ret;
}
EXPORT_SYMBOL_GPL(kvm_emulate_pio_string);

2390
int kvm_arch_init(void *opaque)
2391
{
2392
	int r;
2393 2394 2395 2396
	struct kvm_x86_ops *ops = (struct kvm_x86_ops *)opaque;

	if (kvm_x86_ops) {
		printk(KERN_ERR "kvm: already loaded the other module\n");
2397 2398
		r = -EEXIST;
		goto out;
2399 2400 2401 2402
	}

	if (!ops->cpu_has_kvm_support()) {
		printk(KERN_ERR "kvm: no hardware support\n");
2403 2404
		r = -EOPNOTSUPP;
		goto out;
2405 2406 2407
	}
	if (ops->disabled_by_bios()) {
		printk(KERN_ERR "kvm: disabled by bios\n");
2408 2409
		r = -EOPNOTSUPP;
		goto out;
2410 2411
	}

2412 2413 2414 2415 2416 2417
	r = kvm_mmu_module_init();
	if (r)
		goto out;

	kvm_init_msr_list();

2418
	kvm_x86_ops = ops;
2419
	kvm_mmu_set_nonpresent_ptes(0ull, 0ull);
2420
	return 0;
2421 2422 2423

out:
	return r;
2424
}
2425

2426 2427 2428
void kvm_arch_exit(void)
{
	kvm_x86_ops = NULL;
2429 2430
	kvm_mmu_module_exit();
}
2431

2432 2433 2434
int kvm_emulate_halt(struct kvm_vcpu *vcpu)
{
	++vcpu->stat.halt_exits;
F
Feng (Eric) Liu 已提交
2435
	KVMTRACE_0D(HLT, vcpu, handler);
2436
	if (irqchip_in_kernel(vcpu->kvm)) {
2437
		vcpu->arch.mp_state = KVM_MP_STATE_HALTED;
2438
		up_read(&vcpu->kvm->slots_lock);
2439
		kvm_vcpu_block(vcpu);
2440
		down_read(&vcpu->kvm->slots_lock);
2441
		if (vcpu->arch.mp_state != KVM_MP_STATE_RUNNABLE)
2442 2443 2444 2445 2446 2447 2448 2449 2450
			return -EINTR;
		return 1;
	} else {
		vcpu->run->exit_reason = KVM_EXIT_HLT;
		return 0;
	}
}
EXPORT_SYMBOL_GPL(kvm_emulate_halt);

2451 2452 2453 2454 2455 2456 2457 2458 2459
static inline gpa_t hc_gpa(struct kvm_vcpu *vcpu, unsigned long a0,
			   unsigned long a1)
{
	if (is_long_mode(vcpu))
		return a0;
	else
		return a0 | ((gpa_t)a1 << 32);
}

2460 2461 2462
int kvm_emulate_hypercall(struct kvm_vcpu *vcpu)
{
	unsigned long nr, a0, a1, a2, a3, ret;
2463
	int r = 1;
2464 2465 2466

	kvm_x86_ops->cache_regs(vcpu);

2467 2468 2469 2470 2471
	nr = vcpu->arch.regs[VCPU_REGS_RAX];
	a0 = vcpu->arch.regs[VCPU_REGS_RBX];
	a1 = vcpu->arch.regs[VCPU_REGS_RCX];
	a2 = vcpu->arch.regs[VCPU_REGS_RDX];
	a3 = vcpu->arch.regs[VCPU_REGS_RSI];
2472

F
Feng (Eric) Liu 已提交
2473 2474
	KVMTRACE_1D(VMMCALL, vcpu, (u32)nr, handler);

2475 2476 2477 2478 2479 2480 2481 2482 2483
	if (!is_long_mode(vcpu)) {
		nr &= 0xFFFFFFFF;
		a0 &= 0xFFFFFFFF;
		a1 &= 0xFFFFFFFF;
		a2 &= 0xFFFFFFFF;
		a3 &= 0xFFFFFFFF;
	}

	switch (nr) {
A
Avi Kivity 已提交
2484 2485 2486
	case KVM_HC_VAPIC_POLL_IRQ:
		ret = 0;
		break;
2487 2488 2489
	case KVM_HC_MMU_OP:
		r = kvm_pv_mmu_op(vcpu, a0, hc_gpa(vcpu, a1, a2), &ret);
		break;
2490 2491 2492 2493
	default:
		ret = -KVM_ENOSYS;
		break;
	}
2494
	vcpu->arch.regs[VCPU_REGS_RAX] = ret;
2495
	kvm_x86_ops->decache_regs(vcpu);
A
Amit Shah 已提交
2496
	++vcpu->stat.hypercalls;
2497
	return r;
2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515
}
EXPORT_SYMBOL_GPL(kvm_emulate_hypercall);

int kvm_fix_hypercall(struct kvm_vcpu *vcpu)
{
	char instruction[3];
	int ret = 0;


	/*
	 * Blow out the MMU to ensure that no other VCPU has an active mapping
	 * to ensure that the updated hypercall appears atomically across all
	 * VCPUs.
	 */
	kvm_mmu_zap_all(vcpu->kvm);

	kvm_x86_ops->cache_regs(vcpu);
	kvm_x86_ops->patch_hypercall(vcpu, instruction);
2516
	if (emulator_write_emulated(vcpu->arch.rip, instruction, 3, vcpu)
2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544
	    != X86EMUL_CONTINUE)
		ret = -EFAULT;

	return ret;
}

static u64 mk_cr_64(u64 curr_cr, u32 new_val)
{
	return (curr_cr & ~((1ULL << 32) - 1)) | new_val;
}

void realmode_lgdt(struct kvm_vcpu *vcpu, u16 limit, unsigned long base)
{
	struct descriptor_table dt = { limit, base };

	kvm_x86_ops->set_gdt(vcpu, &dt);
}

void realmode_lidt(struct kvm_vcpu *vcpu, u16 limit, unsigned long base)
{
	struct descriptor_table dt = { limit, base };

	kvm_x86_ops->set_idt(vcpu, &dt);
}

void realmode_lmsw(struct kvm_vcpu *vcpu, unsigned long msw,
		   unsigned long *rflags)
{
2545
	kvm_lmsw(vcpu, msw);
2546 2547 2548 2549 2550 2551 2552 2553
	*rflags = kvm_x86_ops->get_rflags(vcpu);
}

unsigned long realmode_get_cr(struct kvm_vcpu *vcpu, int cr)
{
	kvm_x86_ops->decache_cr4_guest_bits(vcpu);
	switch (cr) {
	case 0:
2554
		return vcpu->arch.cr0;
2555
	case 2:
2556
		return vcpu->arch.cr2;
2557
	case 3:
2558
		return vcpu->arch.cr3;
2559
	case 4:
2560
		return vcpu->arch.cr4;
2561
	case 8:
2562
		return kvm_get_cr8(vcpu);
2563
	default:
2564
		vcpu_printf(vcpu, "%s: unexpected cr %u\n", __func__, cr);
2565 2566 2567 2568 2569 2570 2571 2572 2573
		return 0;
	}
}

void realmode_set_cr(struct kvm_vcpu *vcpu, int cr, unsigned long val,
		     unsigned long *rflags)
{
	switch (cr) {
	case 0:
2574
		kvm_set_cr0(vcpu, mk_cr_64(vcpu->arch.cr0, val));
2575 2576 2577
		*rflags = kvm_x86_ops->get_rflags(vcpu);
		break;
	case 2:
2578
		vcpu->arch.cr2 = val;
2579 2580
		break;
	case 3:
2581
		kvm_set_cr3(vcpu, val);
2582 2583
		break;
	case 4:
2584
		kvm_set_cr4(vcpu, mk_cr_64(vcpu->arch.cr4, val));
2585
		break;
2586
	case 8:
2587
		kvm_set_cr8(vcpu, val & 0xfUL);
2588
		break;
2589
	default:
2590
		vcpu_printf(vcpu, "%s: unexpected cr %u\n", __func__, cr);
2591 2592 2593
	}
}

2594 2595
static int move_to_next_stateful_cpuid_entry(struct kvm_vcpu *vcpu, int i)
{
2596 2597
	struct kvm_cpuid_entry2 *e = &vcpu->arch.cpuid_entries[i];
	int j, nent = vcpu->arch.cpuid_nent;
2598 2599 2600 2601

	e->flags &= ~KVM_CPUID_FLAG_STATE_READ_NEXT;
	/* when no next entry is found, the current entry[i] is reselected */
	for (j = i + 1; j == i; j = (j + 1) % nent) {
2602
		struct kvm_cpuid_entry2 *ej = &vcpu->arch.cpuid_entries[j];
2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625
		if (ej->function == e->function) {
			ej->flags |= KVM_CPUID_FLAG_STATE_READ_NEXT;
			return j;
		}
	}
	return 0; /* silence gcc, even though control never reaches here */
}

/* find an entry with matching function, matching index (if needed), and that
 * should be read next (if it's stateful) */
static int is_matching_cpuid_entry(struct kvm_cpuid_entry2 *e,
	u32 function, u32 index)
{
	if (e->function != function)
		return 0;
	if ((e->flags & KVM_CPUID_FLAG_SIGNIFCANT_INDEX) && e->index != index)
		return 0;
	if ((e->flags & KVM_CPUID_FLAG_STATEFUL_FUNC) &&
		!(e->flags & KVM_CPUID_FLAG_STATE_READ_NEXT))
		return 0;
	return 1;
}

2626 2627 2628
void kvm_emulate_cpuid(struct kvm_vcpu *vcpu)
{
	int i;
2629 2630
	u32 function, index;
	struct kvm_cpuid_entry2 *e, *best;
2631 2632

	kvm_x86_ops->cache_regs(vcpu);
2633 2634 2635 2636 2637 2638
	function = vcpu->arch.regs[VCPU_REGS_RAX];
	index = vcpu->arch.regs[VCPU_REGS_RCX];
	vcpu->arch.regs[VCPU_REGS_RAX] = 0;
	vcpu->arch.regs[VCPU_REGS_RBX] = 0;
	vcpu->arch.regs[VCPU_REGS_RCX] = 0;
	vcpu->arch.regs[VCPU_REGS_RDX] = 0;
2639
	best = NULL;
2640 2641
	for (i = 0; i < vcpu->arch.cpuid_nent; ++i) {
		e = &vcpu->arch.cpuid_entries[i];
2642 2643 2644
		if (is_matching_cpuid_entry(e, function, index)) {
			if (e->flags & KVM_CPUID_FLAG_STATEFUL_FUNC)
				move_to_next_stateful_cpuid_entry(vcpu, i);
2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655
			best = e;
			break;
		}
		/*
		 * Both basic or both extended?
		 */
		if (((e->function ^ function) & 0x80000000) == 0)
			if (!best || e->function > best->function)
				best = e;
	}
	if (best) {
2656 2657 2658 2659
		vcpu->arch.regs[VCPU_REGS_RAX] = best->eax;
		vcpu->arch.regs[VCPU_REGS_RBX] = best->ebx;
		vcpu->arch.regs[VCPU_REGS_RCX] = best->ecx;
		vcpu->arch.regs[VCPU_REGS_RDX] = best->edx;
2660 2661 2662
	}
	kvm_x86_ops->decache_regs(vcpu);
	kvm_x86_ops->skip_emulated_instruction(vcpu);
F
Feng (Eric) Liu 已提交
2663 2664 2665 2666 2667
	KVMTRACE_5D(CPUID, vcpu, function,
		    (u32)vcpu->arch.regs[VCPU_REGS_RAX],
		    (u32)vcpu->arch.regs[VCPU_REGS_RBX],
		    (u32)vcpu->arch.regs[VCPU_REGS_RCX],
		    (u32)vcpu->arch.regs[VCPU_REGS_RDX], handler);
2668 2669
}
EXPORT_SYMBOL_GPL(kvm_emulate_cpuid);
2670

2671 2672 2673 2674 2675 2676 2677 2678 2679
/*
 * Check if userspace requested an interrupt window, and that the
 * interrupt window is open.
 *
 * No need to exit to userspace if we already have an interrupt queued.
 */
static int dm_request_for_irq_injection(struct kvm_vcpu *vcpu,
					  struct kvm_run *kvm_run)
{
2680
	return (!vcpu->arch.irq_summary &&
2681
		kvm_run->request_interrupt_window &&
2682
		vcpu->arch.interrupt_window_open &&
2683 2684 2685 2686 2687 2688 2689
		(kvm_x86_ops->get_rflags(vcpu) & X86_EFLAGS_IF));
}

static void post_kvm_run_save(struct kvm_vcpu *vcpu,
			      struct kvm_run *kvm_run)
{
	kvm_run->if_flag = (kvm_x86_ops->get_rflags(vcpu) & X86_EFLAGS_IF) != 0;
2690
	kvm_run->cr8 = kvm_get_cr8(vcpu);
2691 2692 2693 2694 2695
	kvm_run->apic_base = kvm_get_apic_base(vcpu);
	if (irqchip_in_kernel(vcpu->kvm))
		kvm_run->ready_for_interrupt_injection = 1;
	else
		kvm_run->ready_for_interrupt_injection =
2696 2697
					(vcpu->arch.interrupt_window_open &&
					 vcpu->arch.irq_summary == 0);
2698 2699
}

A
Avi Kivity 已提交
2700 2701 2702 2703 2704 2705 2706 2707
static void vapic_enter(struct kvm_vcpu *vcpu)
{
	struct kvm_lapic *apic = vcpu->arch.apic;
	struct page *page;

	if (!apic || !apic->vapic_addr)
		return;

2708
	down_read(&current->mm->mmap_sem);
A
Avi Kivity 已提交
2709
	page = gfn_to_page(vcpu->kvm, apic->vapic_addr >> PAGE_SHIFT);
2710
	up_read(&current->mm->mmap_sem);
2711 2712

	vcpu->arch.apic->vapic_page = page;
A
Avi Kivity 已提交
2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725
}

static void vapic_exit(struct kvm_vcpu *vcpu)
{
	struct kvm_lapic *apic = vcpu->arch.apic;

	if (!apic || !apic->vapic_addr)
		return;

	kvm_release_page_dirty(apic->vapic_page);
	mark_page_dirty(vcpu->kvm, apic->vapic_addr >> PAGE_SHIFT);
}

2726 2727 2728 2729
static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
{
	int r;

2730
	if (unlikely(vcpu->arch.mp_state == KVM_MP_STATE_SIPI_RECEIVED)) {
2731
		pr_debug("vcpu %d received sipi with vector # %x\n",
2732
		       vcpu->vcpu_id, vcpu->arch.sipi_vector);
2733 2734 2735 2736
		kvm_lapic_reset(vcpu);
		r = kvm_x86_ops->vcpu_reset(vcpu);
		if (r)
			return r;
2737
		vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
2738 2739
	}

2740
	down_read(&vcpu->kvm->slots_lock);
A
Avi Kivity 已提交
2741 2742
	vapic_enter(vcpu);

2743 2744 2745 2746 2747
preempted:
	if (vcpu->guest_debug.enabled)
		kvm_x86_ops->guest_debug_pre(vcpu);

again:
2748 2749 2750 2751
	if (vcpu->requests)
		if (test_and_clear_bit(KVM_REQ_MMU_RELOAD, &vcpu->requests))
			kvm_mmu_unload(vcpu);

2752 2753 2754 2755
	r = kvm_mmu_reload(vcpu);
	if (unlikely(r))
		goto out;

2756 2757 2758
	if (vcpu->requests) {
		if (test_and_clear_bit(KVM_REQ_MIGRATE_TIMER, &vcpu->requests))
			__kvm_migrate_apic_timer(vcpu);
A
Avi Kivity 已提交
2759 2760 2761 2762 2763 2764
		if (test_and_clear_bit(KVM_REQ_REPORT_TPR_ACCESS,
				       &vcpu->requests)) {
			kvm_run->exit_reason = KVM_EXIT_TPR_ACCESS;
			r = 0;
			goto out;
		}
J
Joerg Roedel 已提交
2765 2766 2767 2768 2769
		if (test_and_clear_bit(KVM_REQ_TRIPLE_FAULT, &vcpu->requests)) {
			kvm_run->exit_reason = KVM_EXIT_SHUTDOWN;
			r = 0;
			goto out;
		}
2770
	}
A
Avi Kivity 已提交
2771

2772 2773 2774 2775 2776 2777 2778 2779 2780
	kvm_inject_pending_timer_irqs(vcpu);

	preempt_disable();

	kvm_x86_ops->prepare_guest_switch(vcpu);
	kvm_load_guest_fpu(vcpu);

	local_irq_disable();

2781 2782 2783 2784 2785 2786 2787
	if (need_resched()) {
		local_irq_enable();
		preempt_enable();
		r = 1;
		goto out;
	}

2788 2789 2790 2791 2792 2793 2794 2795
	if (vcpu->requests)
		if (test_bit(KVM_REQ_MMU_RELOAD, &vcpu->requests)) {
			local_irq_enable();
			preempt_enable();
			r = 1;
			goto out;
		}

2796 2797 2798 2799 2800 2801 2802 2803 2804
	if (signal_pending(current)) {
		local_irq_enable();
		preempt_enable();
		r = -EINTR;
		kvm_run->exit_reason = KVM_EXIT_INTR;
		++vcpu->stat.signal_exits;
		goto out;
	}

2805 2806 2807 2808 2809 2810 2811
	vcpu->guest_mode = 1;
	/*
	 * Make sure that guest_mode assignment won't happen after
	 * testing the pending IRQ vector bitmap.
	 */
	smp_wmb();

2812
	if (vcpu->arch.exception.pending)
2813 2814
		__queue_exception(vcpu);
	else if (irqchip_in_kernel(vcpu->kvm))
2815
		kvm_x86_ops->inject_pending_irq(vcpu);
2816
	else
2817 2818
		kvm_x86_ops->inject_pending_vectors(vcpu, kvm_run);

A
Avi Kivity 已提交
2819 2820
	kvm_lapic_sync_to_vapic(vcpu);

2821 2822
	up_read(&vcpu->kvm->slots_lock);

2823 2824 2825 2826 2827 2828
	kvm_guest_enter();

	if (vcpu->requests)
		if (test_and_clear_bit(KVM_REQ_TLB_FLUSH, &vcpu->requests))
			kvm_x86_ops->tlb_flush(vcpu);

F
Feng (Eric) Liu 已提交
2829
	KVMTRACE_0D(VMENTRY, vcpu, entryexit);
2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848
	kvm_x86_ops->run(vcpu, kvm_run);

	vcpu->guest_mode = 0;
	local_irq_enable();

	++vcpu->stat.exits;

	/*
	 * We must have an instruction between local_irq_enable() and
	 * kvm_guest_exit(), so the timer interrupt isn't delayed by
	 * the interrupt shadow.  The stat.exits increment will do nicely.
	 * But we need to prevent reordering, hence this barrier():
	 */
	barrier();

	kvm_guest_exit();

	preempt_enable();

2849 2850
	down_read(&vcpu->kvm->slots_lock);

2851 2852 2853 2854 2855
	/*
	 * Profile KVM exit RIPs:
	 */
	if (unlikely(prof_on == KVM_PROFILING)) {
		kvm_x86_ops->cache_regs(vcpu);
2856
		profile_hit(KVM_PROFILING, (void *)vcpu->arch.rip);
2857 2858
	}

2859 2860
	if (vcpu->arch.exception.pending && kvm_x86_ops->exception_injected(vcpu))
		vcpu->arch.exception.pending = false;
2861

A
Avi Kivity 已提交
2862 2863
	kvm_lapic_sync_from_vapic(vcpu);

2864 2865 2866 2867 2868 2869 2870 2871 2872
	r = kvm_x86_ops->handle_exit(kvm_run, vcpu);

	if (r > 0) {
		if (dm_request_for_irq_injection(vcpu, kvm_run)) {
			r = -EINTR;
			kvm_run->exit_reason = KVM_EXIT_INTR;
			++vcpu->stat.request_irq_exits;
			goto out;
		}
2873
		if (!need_resched())
2874 2875 2876 2877
			goto again;
	}

out:
2878
	up_read(&vcpu->kvm->slots_lock);
2879 2880
	if (r > 0) {
		kvm_resched(vcpu);
2881
		down_read(&vcpu->kvm->slots_lock);
2882 2883 2884 2885 2886
		goto preempted;
	}

	post_kvm_run_save(vcpu, kvm_run);

2887
	down_read(&vcpu->kvm->slots_lock);
A
Avi Kivity 已提交
2888
	vapic_exit(vcpu);
2889
	up_read(&vcpu->kvm->slots_lock);
A
Avi Kivity 已提交
2890

2891 2892 2893 2894 2895 2896 2897 2898 2899 2900
	return r;
}

int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
{
	int r;
	sigset_t sigsaved;

	vcpu_load(vcpu);

2901
	if (unlikely(vcpu->arch.mp_state == KVM_MP_STATE_UNINITIALIZED)) {
2902 2903 2904 2905 2906 2907 2908 2909 2910 2911
		kvm_vcpu_block(vcpu);
		vcpu_put(vcpu);
		return -EAGAIN;
	}

	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);

	/* re-sync apic's tpr */
	if (!irqchip_in_kernel(vcpu->kvm))
2912
		kvm_set_cr8(vcpu, kvm_run->cr8);
2913

2914
	if (vcpu->arch.pio.cur_count) {
2915 2916 2917 2918 2919 2920 2921 2922 2923
		r = complete_pio(vcpu);
		if (r)
			goto out;
	}
#if CONFIG_HAS_IOMEM
	if (vcpu->mmio_needed) {
		memcpy(vcpu->mmio_data, kvm_run->mmio.data, 8);
		vcpu->mmio_read_completed = 1;
		vcpu->mmio_needed = 0;
2924 2925

		down_read(&vcpu->kvm->slots_lock);
2926
		r = emulate_instruction(vcpu, kvm_run,
2927 2928
					vcpu->arch.mmio_fault_cr2, 0,
					EMULTYPE_NO_DECODE);
2929
		up_read(&vcpu->kvm->slots_lock);
2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940
		if (r == EMULATE_DO_MMIO) {
			/*
			 * Read-modify-write.  Back to userspace.
			 */
			r = 0;
			goto out;
		}
	}
#endif
	if (kvm_run->exit_reason == KVM_EXIT_HYPERCALL) {
		kvm_x86_ops->cache_regs(vcpu);
2941
		vcpu->arch.regs[VCPU_REGS_RAX] = kvm_run->hypercall.ret;
2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960
		kvm_x86_ops->decache_regs(vcpu);
	}

	r = __vcpu_run(vcpu, kvm_run);

out:
	if (vcpu->sigset_active)
		sigprocmask(SIG_SETMASK, &sigsaved, NULL);

	vcpu_put(vcpu);
	return r;
}

int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
	vcpu_load(vcpu);

	kvm_x86_ops->cache_regs(vcpu);

2961 2962 2963 2964 2965 2966 2967 2968
	regs->rax = vcpu->arch.regs[VCPU_REGS_RAX];
	regs->rbx = vcpu->arch.regs[VCPU_REGS_RBX];
	regs->rcx = vcpu->arch.regs[VCPU_REGS_RCX];
	regs->rdx = vcpu->arch.regs[VCPU_REGS_RDX];
	regs->rsi = vcpu->arch.regs[VCPU_REGS_RSI];
	regs->rdi = vcpu->arch.regs[VCPU_REGS_RDI];
	regs->rsp = vcpu->arch.regs[VCPU_REGS_RSP];
	regs->rbp = vcpu->arch.regs[VCPU_REGS_RBP];
2969
#ifdef CONFIG_X86_64
2970 2971 2972 2973 2974 2975 2976 2977
	regs->r8 = vcpu->arch.regs[VCPU_REGS_R8];
	regs->r9 = vcpu->arch.regs[VCPU_REGS_R9];
	regs->r10 = vcpu->arch.regs[VCPU_REGS_R10];
	regs->r11 = vcpu->arch.regs[VCPU_REGS_R11];
	regs->r12 = vcpu->arch.regs[VCPU_REGS_R12];
	regs->r13 = vcpu->arch.regs[VCPU_REGS_R13];
	regs->r14 = vcpu->arch.regs[VCPU_REGS_R14];
	regs->r15 = vcpu->arch.regs[VCPU_REGS_R15];
2978 2979
#endif

2980
	regs->rip = vcpu->arch.rip;
2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997
	regs->rflags = kvm_x86_ops->get_rflags(vcpu);

	/*
	 * Don't leak debug flags in case they were set for guest debugging
	 */
	if (vcpu->guest_debug.enabled && vcpu->guest_debug.singlestep)
		regs->rflags &= ~(X86_EFLAGS_TF | X86_EFLAGS_RF);

	vcpu_put(vcpu);

	return 0;
}

int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
	vcpu_load(vcpu);

2998 2999 3000 3001 3002 3003 3004 3005
	vcpu->arch.regs[VCPU_REGS_RAX] = regs->rax;
	vcpu->arch.regs[VCPU_REGS_RBX] = regs->rbx;
	vcpu->arch.regs[VCPU_REGS_RCX] = regs->rcx;
	vcpu->arch.regs[VCPU_REGS_RDX] = regs->rdx;
	vcpu->arch.regs[VCPU_REGS_RSI] = regs->rsi;
	vcpu->arch.regs[VCPU_REGS_RDI] = regs->rdi;
	vcpu->arch.regs[VCPU_REGS_RSP] = regs->rsp;
	vcpu->arch.regs[VCPU_REGS_RBP] = regs->rbp;
3006
#ifdef CONFIG_X86_64
3007 3008 3009 3010 3011 3012 3013 3014
	vcpu->arch.regs[VCPU_REGS_R8] = regs->r8;
	vcpu->arch.regs[VCPU_REGS_R9] = regs->r9;
	vcpu->arch.regs[VCPU_REGS_R10] = regs->r10;
	vcpu->arch.regs[VCPU_REGS_R11] = regs->r11;
	vcpu->arch.regs[VCPU_REGS_R12] = regs->r12;
	vcpu->arch.regs[VCPU_REGS_R13] = regs->r13;
	vcpu->arch.regs[VCPU_REGS_R14] = regs->r14;
	vcpu->arch.regs[VCPU_REGS_R15] = regs->r15;
3015 3016
#endif

3017
	vcpu->arch.rip = regs->rip;
3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029
	kvm_x86_ops->set_rflags(vcpu, regs->rflags);

	kvm_x86_ops->decache_regs(vcpu);

	vcpu_put(vcpu);

	return 0;
}

static void get_segment(struct kvm_vcpu *vcpu,
			struct kvm_segment *var, int seg)
{
3030
	kvm_x86_ops->get_segment(vcpu, var, seg);
3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068
}

void kvm_get_cs_db_l_bits(struct kvm_vcpu *vcpu, int *db, int *l)
{
	struct kvm_segment cs;

	get_segment(vcpu, &cs, VCPU_SREG_CS);
	*db = cs.db;
	*l = cs.l;
}
EXPORT_SYMBOL_GPL(kvm_get_cs_db_l_bits);

int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs)
{
	struct descriptor_table dt;
	int pending_vec;

	vcpu_load(vcpu);

	get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
	get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
	get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
	get_segment(vcpu, &sregs->fs, VCPU_SREG_FS);
	get_segment(vcpu, &sregs->gs, VCPU_SREG_GS);
	get_segment(vcpu, &sregs->ss, VCPU_SREG_SS);

	get_segment(vcpu, &sregs->tr, VCPU_SREG_TR);
	get_segment(vcpu, &sregs->ldt, VCPU_SREG_LDTR);

	kvm_x86_ops->get_idt(vcpu, &dt);
	sregs->idt.limit = dt.limit;
	sregs->idt.base = dt.base;
	kvm_x86_ops->get_gdt(vcpu, &dt);
	sregs->gdt.limit = dt.limit;
	sregs->gdt.base = dt.base;

	kvm_x86_ops->decache_cr4_guest_bits(vcpu);
3069 3070 3071 3072
	sregs->cr0 = vcpu->arch.cr0;
	sregs->cr2 = vcpu->arch.cr2;
	sregs->cr3 = vcpu->arch.cr3;
	sregs->cr4 = vcpu->arch.cr4;
3073
	sregs->cr8 = kvm_get_cr8(vcpu);
3074
	sregs->efer = vcpu->arch.shadow_efer;
3075 3076 3077 3078 3079 3080 3081 3082 3083 3084
	sregs->apic_base = kvm_get_apic_base(vcpu);

	if (irqchip_in_kernel(vcpu->kvm)) {
		memset(sregs->interrupt_bitmap, 0,
		       sizeof sregs->interrupt_bitmap);
		pending_vec = kvm_x86_ops->get_irq(vcpu);
		if (pending_vec >= 0)
			set_bit(pending_vec,
				(unsigned long *)sregs->interrupt_bitmap);
	} else
3085
		memcpy(sregs->interrupt_bitmap, vcpu->arch.irq_pending,
3086 3087 3088 3089 3090 3091 3092
		       sizeof sregs->interrupt_bitmap);

	vcpu_put(vcpu);

	return 0;
}

3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110
int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
				    struct kvm_mp_state *mp_state)
{
	vcpu_load(vcpu);
	mp_state->mp_state = vcpu->arch.mp_state;
	vcpu_put(vcpu);
	return 0;
}

int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
				    struct kvm_mp_state *mp_state)
{
	vcpu_load(vcpu);
	vcpu->arch.mp_state = mp_state->mp_state;
	vcpu_put(vcpu);
	return 0;
}

3111 3112 3113
static void set_segment(struct kvm_vcpu *vcpu,
			struct kvm_segment *var, int seg)
{
3114
	kvm_x86_ops->set_segment(vcpu, var, seg);
3115 3116
}

3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 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 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524
static void seg_desct_to_kvm_desct(struct desc_struct *seg_desc, u16 selector,
				   struct kvm_segment *kvm_desct)
{
	kvm_desct->base = seg_desc->base0;
	kvm_desct->base |= seg_desc->base1 << 16;
	kvm_desct->base |= seg_desc->base2 << 24;
	kvm_desct->limit = seg_desc->limit0;
	kvm_desct->limit |= seg_desc->limit << 16;
	kvm_desct->selector = selector;
	kvm_desct->type = seg_desc->type;
	kvm_desct->present = seg_desc->p;
	kvm_desct->dpl = seg_desc->dpl;
	kvm_desct->db = seg_desc->d;
	kvm_desct->s = seg_desc->s;
	kvm_desct->l = seg_desc->l;
	kvm_desct->g = seg_desc->g;
	kvm_desct->avl = seg_desc->avl;
	if (!selector)
		kvm_desct->unusable = 1;
	else
		kvm_desct->unusable = 0;
	kvm_desct->padding = 0;
}

static void get_segment_descritptor_dtable(struct kvm_vcpu *vcpu,
					   u16 selector,
					   struct descriptor_table *dtable)
{
	if (selector & 1 << 2) {
		struct kvm_segment kvm_seg;

		get_segment(vcpu, &kvm_seg, VCPU_SREG_LDTR);

		if (kvm_seg.unusable)
			dtable->limit = 0;
		else
			dtable->limit = kvm_seg.limit;
		dtable->base = kvm_seg.base;
	}
	else
		kvm_x86_ops->get_gdt(vcpu, dtable);
}

/* allowed just for 8 bytes segments */
static int load_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
					 struct desc_struct *seg_desc)
{
	struct descriptor_table dtable;
	u16 index = selector >> 3;

	get_segment_descritptor_dtable(vcpu, selector, &dtable);

	if (dtable.limit < index * 8 + 7) {
		kvm_queue_exception_e(vcpu, GP_VECTOR, selector & 0xfffc);
		return 1;
	}
	return kvm_read_guest(vcpu->kvm, dtable.base + index * 8, seg_desc, 8);
}

/* allowed just for 8 bytes segments */
static int save_guest_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
					 struct desc_struct *seg_desc)
{
	struct descriptor_table dtable;
	u16 index = selector >> 3;

	get_segment_descritptor_dtable(vcpu, selector, &dtable);

	if (dtable.limit < index * 8 + 7)
		return 1;
	return kvm_write_guest(vcpu->kvm, dtable.base + index * 8, seg_desc, 8);
}

static u32 get_tss_base_addr(struct kvm_vcpu *vcpu,
			     struct desc_struct *seg_desc)
{
	u32 base_addr;

	base_addr = seg_desc->base0;
	base_addr |= (seg_desc->base1 << 16);
	base_addr |= (seg_desc->base2 << 24);

	return base_addr;
}

static int load_tss_segment32(struct kvm_vcpu *vcpu,
			      struct desc_struct *seg_desc,
			      struct tss_segment_32 *tss)
{
	u32 base_addr;

	base_addr = get_tss_base_addr(vcpu, seg_desc);

	return kvm_read_guest(vcpu->kvm, base_addr, tss,
			      sizeof(struct tss_segment_32));
}

static int save_tss_segment32(struct kvm_vcpu *vcpu,
			      struct desc_struct *seg_desc,
			      struct tss_segment_32 *tss)
{
	u32 base_addr;

	base_addr = get_tss_base_addr(vcpu, seg_desc);

	return kvm_write_guest(vcpu->kvm, base_addr, tss,
			       sizeof(struct tss_segment_32));
}

static int load_tss_segment16(struct kvm_vcpu *vcpu,
			      struct desc_struct *seg_desc,
			      struct tss_segment_16 *tss)
{
	u32 base_addr;

	base_addr = get_tss_base_addr(vcpu, seg_desc);

	return kvm_read_guest(vcpu->kvm, base_addr, tss,
			      sizeof(struct tss_segment_16));
}

static int save_tss_segment16(struct kvm_vcpu *vcpu,
			      struct desc_struct *seg_desc,
			      struct tss_segment_16 *tss)
{
	u32 base_addr;

	base_addr = get_tss_base_addr(vcpu, seg_desc);

	return kvm_write_guest(vcpu->kvm, base_addr, tss,
			       sizeof(struct tss_segment_16));
}

static u16 get_segment_selector(struct kvm_vcpu *vcpu, int seg)
{
	struct kvm_segment kvm_seg;

	get_segment(vcpu, &kvm_seg, seg);
	return kvm_seg.selector;
}

static int load_segment_descriptor_to_kvm_desct(struct kvm_vcpu *vcpu,
						u16 selector,
						struct kvm_segment *kvm_seg)
{
	struct desc_struct seg_desc;

	if (load_guest_segment_descriptor(vcpu, selector, &seg_desc))
		return 1;
	seg_desct_to_kvm_desct(&seg_desc, selector, kvm_seg);
	return 0;
}

static int load_segment_descriptor(struct kvm_vcpu *vcpu, u16 selector,
				   int type_bits, int seg)
{
	struct kvm_segment kvm_seg;

	if (load_segment_descriptor_to_kvm_desct(vcpu, selector, &kvm_seg))
		return 1;
	kvm_seg.type |= type_bits;

	if (seg != VCPU_SREG_SS && seg != VCPU_SREG_CS &&
	    seg != VCPU_SREG_LDTR)
		if (!kvm_seg.s)
			kvm_seg.unusable = 1;

	set_segment(vcpu, &kvm_seg, seg);
	return 0;
}

static void save_state_to_tss32(struct kvm_vcpu *vcpu,
				struct tss_segment_32 *tss)
{
	tss->cr3 = vcpu->arch.cr3;
	tss->eip = vcpu->arch.rip;
	tss->eflags = kvm_x86_ops->get_rflags(vcpu);
	tss->eax = vcpu->arch.regs[VCPU_REGS_RAX];
	tss->ecx = vcpu->arch.regs[VCPU_REGS_RCX];
	tss->edx = vcpu->arch.regs[VCPU_REGS_RDX];
	tss->ebx = vcpu->arch.regs[VCPU_REGS_RBX];
	tss->esp = vcpu->arch.regs[VCPU_REGS_RSP];
	tss->ebp = vcpu->arch.regs[VCPU_REGS_RBP];
	tss->esi = vcpu->arch.regs[VCPU_REGS_RSI];
	tss->edi = vcpu->arch.regs[VCPU_REGS_RDI];

	tss->es = get_segment_selector(vcpu, VCPU_SREG_ES);
	tss->cs = get_segment_selector(vcpu, VCPU_SREG_CS);
	tss->ss = get_segment_selector(vcpu, VCPU_SREG_SS);
	tss->ds = get_segment_selector(vcpu, VCPU_SREG_DS);
	tss->fs = get_segment_selector(vcpu, VCPU_SREG_FS);
	tss->gs = get_segment_selector(vcpu, VCPU_SREG_GS);
	tss->ldt_selector = get_segment_selector(vcpu, VCPU_SREG_LDTR);
	tss->prev_task_link = get_segment_selector(vcpu, VCPU_SREG_TR);
}

static int load_state_from_tss32(struct kvm_vcpu *vcpu,
				  struct tss_segment_32 *tss)
{
	kvm_set_cr3(vcpu, tss->cr3);

	vcpu->arch.rip = tss->eip;
	kvm_x86_ops->set_rflags(vcpu, tss->eflags | 2);

	vcpu->arch.regs[VCPU_REGS_RAX] = tss->eax;
	vcpu->arch.regs[VCPU_REGS_RCX] = tss->ecx;
	vcpu->arch.regs[VCPU_REGS_RDX] = tss->edx;
	vcpu->arch.regs[VCPU_REGS_RBX] = tss->ebx;
	vcpu->arch.regs[VCPU_REGS_RSP] = tss->esp;
	vcpu->arch.regs[VCPU_REGS_RBP] = tss->ebp;
	vcpu->arch.regs[VCPU_REGS_RSI] = tss->esi;
	vcpu->arch.regs[VCPU_REGS_RDI] = tss->edi;

	if (load_segment_descriptor(vcpu, tss->ldt_selector, 0, VCPU_SREG_LDTR))
		return 1;

	if (load_segment_descriptor(vcpu, tss->es, 1, VCPU_SREG_ES))
		return 1;

	if (load_segment_descriptor(vcpu, tss->cs, 9, VCPU_SREG_CS))
		return 1;

	if (load_segment_descriptor(vcpu, tss->ss, 1, VCPU_SREG_SS))
		return 1;

	if (load_segment_descriptor(vcpu, tss->ds, 1, VCPU_SREG_DS))
		return 1;

	if (load_segment_descriptor(vcpu, tss->fs, 1, VCPU_SREG_FS))
		return 1;

	if (load_segment_descriptor(vcpu, tss->gs, 1, VCPU_SREG_GS))
		return 1;
	return 0;
}

static void save_state_to_tss16(struct kvm_vcpu *vcpu,
				struct tss_segment_16 *tss)
{
	tss->ip = vcpu->arch.rip;
	tss->flag = kvm_x86_ops->get_rflags(vcpu);
	tss->ax = vcpu->arch.regs[VCPU_REGS_RAX];
	tss->cx = vcpu->arch.regs[VCPU_REGS_RCX];
	tss->dx = vcpu->arch.regs[VCPU_REGS_RDX];
	tss->bx = vcpu->arch.regs[VCPU_REGS_RBX];
	tss->sp = vcpu->arch.regs[VCPU_REGS_RSP];
	tss->bp = vcpu->arch.regs[VCPU_REGS_RBP];
	tss->si = vcpu->arch.regs[VCPU_REGS_RSI];
	tss->di = vcpu->arch.regs[VCPU_REGS_RDI];

	tss->es = get_segment_selector(vcpu, VCPU_SREG_ES);
	tss->cs = get_segment_selector(vcpu, VCPU_SREG_CS);
	tss->ss = get_segment_selector(vcpu, VCPU_SREG_SS);
	tss->ds = get_segment_selector(vcpu, VCPU_SREG_DS);
	tss->ldt = get_segment_selector(vcpu, VCPU_SREG_LDTR);
	tss->prev_task_link = get_segment_selector(vcpu, VCPU_SREG_TR);
}

static int load_state_from_tss16(struct kvm_vcpu *vcpu,
				 struct tss_segment_16 *tss)
{
	vcpu->arch.rip = tss->ip;
	kvm_x86_ops->set_rflags(vcpu, tss->flag | 2);
	vcpu->arch.regs[VCPU_REGS_RAX] = tss->ax;
	vcpu->arch.regs[VCPU_REGS_RCX] = tss->cx;
	vcpu->arch.regs[VCPU_REGS_RDX] = tss->dx;
	vcpu->arch.regs[VCPU_REGS_RBX] = tss->bx;
	vcpu->arch.regs[VCPU_REGS_RSP] = tss->sp;
	vcpu->arch.regs[VCPU_REGS_RBP] = tss->bp;
	vcpu->arch.regs[VCPU_REGS_RSI] = tss->si;
	vcpu->arch.regs[VCPU_REGS_RDI] = tss->di;

	if (load_segment_descriptor(vcpu, tss->ldt, 0, VCPU_SREG_LDTR))
		return 1;

	if (load_segment_descriptor(vcpu, tss->es, 1, VCPU_SREG_ES))
		return 1;

	if (load_segment_descriptor(vcpu, tss->cs, 9, VCPU_SREG_CS))
		return 1;

	if (load_segment_descriptor(vcpu, tss->ss, 1, VCPU_SREG_SS))
		return 1;

	if (load_segment_descriptor(vcpu, tss->ds, 1, VCPU_SREG_DS))
		return 1;
	return 0;
}

int kvm_task_switch_16(struct kvm_vcpu *vcpu, u16 tss_selector,
		       struct desc_struct *cseg_desc,
		       struct desc_struct *nseg_desc)
{
	struct tss_segment_16 tss_segment_16;
	int ret = 0;

	if (load_tss_segment16(vcpu, cseg_desc, &tss_segment_16))
		goto out;

	save_state_to_tss16(vcpu, &tss_segment_16);
	save_tss_segment16(vcpu, cseg_desc, &tss_segment_16);

	if (load_tss_segment16(vcpu, nseg_desc, &tss_segment_16))
		goto out;
	if (load_state_from_tss16(vcpu, &tss_segment_16))
		goto out;

	ret = 1;
out:
	return ret;
}

int kvm_task_switch_32(struct kvm_vcpu *vcpu, u16 tss_selector,
		       struct desc_struct *cseg_desc,
		       struct desc_struct *nseg_desc)
{
	struct tss_segment_32 tss_segment_32;
	int ret = 0;

	if (load_tss_segment32(vcpu, cseg_desc, &tss_segment_32))
		goto out;

	save_state_to_tss32(vcpu, &tss_segment_32);
	save_tss_segment32(vcpu, cseg_desc, &tss_segment_32);

	if (load_tss_segment32(vcpu, nseg_desc, &tss_segment_32))
		goto out;
	if (load_state_from_tss32(vcpu, &tss_segment_32))
		goto out;

	ret = 1;
out:
	return ret;
}

int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int reason)
{
	struct kvm_segment tr_seg;
	struct desc_struct cseg_desc;
	struct desc_struct nseg_desc;
	int ret = 0;

	get_segment(vcpu, &tr_seg, VCPU_SREG_TR);

	if (load_guest_segment_descriptor(vcpu, tss_selector, &nseg_desc))
		goto out;

	if (load_guest_segment_descriptor(vcpu, tr_seg.selector, &cseg_desc))
		goto out;


	if (reason != TASK_SWITCH_IRET) {
		int cpl;

		cpl = kvm_x86_ops->get_cpl(vcpu);
		if ((tss_selector & 3) > nseg_desc.dpl || cpl > nseg_desc.dpl) {
			kvm_queue_exception_e(vcpu, GP_VECTOR, 0);
			return 1;
		}
	}

	if (!nseg_desc.p || (nseg_desc.limit0 | nseg_desc.limit << 16) < 0x67) {
		kvm_queue_exception_e(vcpu, TS_VECTOR, tss_selector & 0xfffc);
		return 1;
	}

	if (reason == TASK_SWITCH_IRET || reason == TASK_SWITCH_JMP) {
		cseg_desc.type &= ~(1 << 8); //clear the B flag
		save_guest_segment_descriptor(vcpu, tr_seg.selector,
					      &cseg_desc);
	}

	if (reason == TASK_SWITCH_IRET) {
		u32 eflags = kvm_x86_ops->get_rflags(vcpu);
		kvm_x86_ops->set_rflags(vcpu, eflags & ~X86_EFLAGS_NT);
	}

	kvm_x86_ops->skip_emulated_instruction(vcpu);
	kvm_x86_ops->cache_regs(vcpu);

	if (nseg_desc.type & 8)
		ret = kvm_task_switch_32(vcpu, tss_selector, &cseg_desc,
					 &nseg_desc);
	else
		ret = kvm_task_switch_16(vcpu, tss_selector, &cseg_desc,
					 &nseg_desc);

	if (reason == TASK_SWITCH_CALL || reason == TASK_SWITCH_GATE) {
		u32 eflags = kvm_x86_ops->get_rflags(vcpu);
		kvm_x86_ops->set_rflags(vcpu, eflags | X86_EFLAGS_NT);
	}

	if (reason != TASK_SWITCH_IRET) {
		nseg_desc.type |= (1 << 8);
		save_guest_segment_descriptor(vcpu, tss_selector,
					      &nseg_desc);
	}

	kvm_x86_ops->set_cr0(vcpu, vcpu->arch.cr0 | X86_CR0_TS);
	seg_desct_to_kvm_desct(&nseg_desc, tss_selector, &tr_seg);
	tr_seg.type = 11;
	set_segment(vcpu, &tr_seg, VCPU_SREG_TR);
out:
	kvm_x86_ops->decache_regs(vcpu);
	return ret;
}
EXPORT_SYMBOL_GPL(kvm_task_switch);

3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540
int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs)
{
	int mmu_reset_needed = 0;
	int i, pending_vec, max_bits;
	struct descriptor_table dt;

	vcpu_load(vcpu);

	dt.limit = sregs->idt.limit;
	dt.base = sregs->idt.base;
	kvm_x86_ops->set_idt(vcpu, &dt);
	dt.limit = sregs->gdt.limit;
	dt.base = sregs->gdt.base;
	kvm_x86_ops->set_gdt(vcpu, &dt);

3541 3542 3543
	vcpu->arch.cr2 = sregs->cr2;
	mmu_reset_needed |= vcpu->arch.cr3 != sregs->cr3;
	vcpu->arch.cr3 = sregs->cr3;
3544

3545
	kvm_set_cr8(vcpu, sregs->cr8);
3546

3547
	mmu_reset_needed |= vcpu->arch.shadow_efer != sregs->efer;
3548 3549 3550 3551 3552
	kvm_x86_ops->set_efer(vcpu, sregs->efer);
	kvm_set_apic_base(vcpu, sregs->apic_base);

	kvm_x86_ops->decache_cr4_guest_bits(vcpu);

3553
	mmu_reset_needed |= vcpu->arch.cr0 != sregs->cr0;
3554
	kvm_x86_ops->set_cr0(vcpu, sregs->cr0);
3555
	vcpu->arch.cr0 = sregs->cr0;
3556

3557
	mmu_reset_needed |= vcpu->arch.cr4 != sregs->cr4;
3558 3559
	kvm_x86_ops->set_cr4(vcpu, sregs->cr4);
	if (!is_long_mode(vcpu) && is_pae(vcpu))
3560
		load_pdptrs(vcpu, vcpu->arch.cr3);
3561 3562 3563 3564 3565

	if (mmu_reset_needed)
		kvm_mmu_reset_context(vcpu);

	if (!irqchip_in_kernel(vcpu->kvm)) {
3566 3567 3568 3569 3570 3571
		memcpy(vcpu->arch.irq_pending, sregs->interrupt_bitmap,
		       sizeof vcpu->arch.irq_pending);
		vcpu->arch.irq_summary = 0;
		for (i = 0; i < ARRAY_SIZE(vcpu->arch.irq_pending); ++i)
			if (vcpu->arch.irq_pending[i])
				__set_bit(i, &vcpu->arch.irq_summary);
3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613
	} else {
		max_bits = (sizeof sregs->interrupt_bitmap) << 3;
		pending_vec = find_first_bit(
			(const unsigned long *)sregs->interrupt_bitmap,
			max_bits);
		/* Only pending external irq is handled here */
		if (pending_vec < max_bits) {
			kvm_x86_ops->set_irq(vcpu, pending_vec);
			pr_debug("Set back pending irq %d\n",
				 pending_vec);
		}
	}

	set_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
	set_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
	set_segment(vcpu, &sregs->es, VCPU_SREG_ES);
	set_segment(vcpu, &sregs->fs, VCPU_SREG_FS);
	set_segment(vcpu, &sregs->gs, VCPU_SREG_GS);
	set_segment(vcpu, &sregs->ss, VCPU_SREG_SS);

	set_segment(vcpu, &sregs->tr, VCPU_SREG_TR);
	set_segment(vcpu, &sregs->ldt, VCPU_SREG_LDTR);

	vcpu_put(vcpu);

	return 0;
}

int kvm_arch_vcpu_ioctl_debug_guest(struct kvm_vcpu *vcpu,
				    struct kvm_debug_guest *dbg)
{
	int r;

	vcpu_load(vcpu);

	r = kvm_x86_ops->set_guest_debug(vcpu, dbg);

	vcpu_put(vcpu);

	return r;
}

3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634
/*
 * fxsave fpu state.  Taken from x86_64/processor.h.  To be killed when
 * we have asm/x86/processor.h
 */
struct fxsave {
	u16	cwd;
	u16	swd;
	u16	twd;
	u16	fop;
	u64	rip;
	u64	rdp;
	u32	mxcsr;
	u32	mxcsr_mask;
	u32	st_space[32];	/* 8*16 bytes for each FP-reg = 128 bytes */
#ifdef CONFIG_X86_64
	u32	xmm_space[64];	/* 16*16 bytes for each XMM-reg = 256 bytes */
#else
	u32	xmm_space[32];	/* 8*16 bytes for each XMM-reg = 128 bytes */
#endif
};

3635 3636 3637 3638 3639 3640 3641 3642 3643 3644
/*
 * Translate a guest virtual address to a guest physical address.
 */
int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				    struct kvm_translation *tr)
{
	unsigned long vaddr = tr->linear_address;
	gpa_t gpa;

	vcpu_load(vcpu);
3645
	down_read(&vcpu->kvm->slots_lock);
3646
	gpa = vcpu->arch.mmu.gva_to_gpa(vcpu, vaddr);
3647
	up_read(&vcpu->kvm->slots_lock);
3648 3649 3650 3651 3652 3653 3654 3655 3656
	tr->physical_address = gpa;
	tr->valid = gpa != UNMAPPED_GVA;
	tr->writeable = 1;
	tr->usermode = 0;
	vcpu_put(vcpu);

	return 0;
}

3657 3658
int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
3659
	struct fxsave *fxsave = (struct fxsave *)&vcpu->arch.guest_fx_image;
3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678

	vcpu_load(vcpu);

	memcpy(fpu->fpr, fxsave->st_space, 128);
	fpu->fcw = fxsave->cwd;
	fpu->fsw = fxsave->swd;
	fpu->ftwx = fxsave->twd;
	fpu->last_opcode = fxsave->fop;
	fpu->last_ip = fxsave->rip;
	fpu->last_dp = fxsave->rdp;
	memcpy(fpu->xmm, fxsave->xmm_space, sizeof fxsave->xmm_space);

	vcpu_put(vcpu);

	return 0;
}

int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
{
3679
	struct fxsave *fxsave = (struct fxsave *)&vcpu->arch.guest_fx_image;
3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702

	vcpu_load(vcpu);

	memcpy(fxsave->st_space, fpu->fpr, 128);
	fxsave->cwd = fpu->fcw;
	fxsave->swd = fpu->fsw;
	fxsave->twd = fpu->ftwx;
	fxsave->fop = fpu->last_opcode;
	fxsave->rip = fpu->last_ip;
	fxsave->rdp = fpu->last_dp;
	memcpy(fxsave->xmm_space, fpu->xmm, sizeof fxsave->xmm_space);

	vcpu_put(vcpu);

	return 0;
}

void fx_init(struct kvm_vcpu *vcpu)
{
	unsigned after_mxcsr_mask;

	/* Initialize guest FPU by resetting ours and saving into guest's */
	preempt_disable();
3703
	fx_save(&vcpu->arch.host_fx_image);
3704
	fpu_init();
3705 3706
	fx_save(&vcpu->arch.guest_fx_image);
	fx_restore(&vcpu->arch.host_fx_image);
3707 3708
	preempt_enable();

3709
	vcpu->arch.cr0 |= X86_CR0_ET;
3710
	after_mxcsr_mask = offsetof(struct i387_fxsave_struct, st_space);
3711 3712
	vcpu->arch.guest_fx_image.mxcsr = 0x1f80;
	memset((void *)&vcpu->arch.guest_fx_image + after_mxcsr_mask,
3713 3714 3715 3716 3717 3718 3719 3720 3721 3722
	       0, sizeof(struct i387_fxsave_struct) - after_mxcsr_mask);
}
EXPORT_SYMBOL_GPL(fx_init);

void kvm_load_guest_fpu(struct kvm_vcpu *vcpu)
{
	if (!vcpu->fpu_active || vcpu->guest_fpu_loaded)
		return;

	vcpu->guest_fpu_loaded = 1;
3723 3724
	fx_save(&vcpu->arch.host_fx_image);
	fx_restore(&vcpu->arch.guest_fx_image);
3725 3726 3727 3728 3729 3730 3731 3732 3733
}
EXPORT_SYMBOL_GPL(kvm_load_guest_fpu);

void kvm_put_guest_fpu(struct kvm_vcpu *vcpu)
{
	if (!vcpu->guest_fpu_loaded)
		return;

	vcpu->guest_fpu_loaded = 0;
3734 3735
	fx_save(&vcpu->arch.guest_fx_image);
	fx_restore(&vcpu->arch.host_fx_image);
A
Avi Kivity 已提交
3736
	++vcpu->stat.fpu_reload;
3737 3738
}
EXPORT_SYMBOL_GPL(kvm_put_guest_fpu);
3739 3740 3741 3742 3743 3744 3745 3746 3747

void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
{
	kvm_x86_ops->vcpu_free(vcpu);
}

struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
						unsigned int id)
{
3748 3749
	return kvm_x86_ops->vcpu_create(kvm, id);
}
3750

3751 3752 3753
int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
{
	int r;
3754 3755

	/* We do fxsave: this must be aligned. */
3756
	BUG_ON((unsigned long)&vcpu->arch.host_fx_image & 0xF);
3757 3758 3759 3760 3761 3762 3763 3764 3765

	vcpu_load(vcpu);
	r = kvm_arch_vcpu_reset(vcpu);
	if (r == 0)
		r = kvm_mmu_setup(vcpu);
	vcpu_put(vcpu);
	if (r < 0)
		goto free_vcpu;

3766
	return 0;
3767 3768
free_vcpu:
	kvm_x86_ops->vcpu_free(vcpu);
3769
	return r;
3770 3771
}

3772
void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819
{
	vcpu_load(vcpu);
	kvm_mmu_unload(vcpu);
	vcpu_put(vcpu);

	kvm_x86_ops->vcpu_free(vcpu);
}

int kvm_arch_vcpu_reset(struct kvm_vcpu *vcpu)
{
	return kvm_x86_ops->vcpu_reset(vcpu);
}

void kvm_arch_hardware_enable(void *garbage)
{
	kvm_x86_ops->hardware_enable(garbage);
}

void kvm_arch_hardware_disable(void *garbage)
{
	kvm_x86_ops->hardware_disable(garbage);
}

int kvm_arch_hardware_setup(void)
{
	return kvm_x86_ops->hardware_setup();
}

void kvm_arch_hardware_unsetup(void)
{
	kvm_x86_ops->hardware_unsetup();
}

void kvm_arch_check_processor_compat(void *rtn)
{
	kvm_x86_ops->check_processor_compatibility(rtn);
}

int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu)
{
	struct page *page;
	struct kvm *kvm;
	int r;

	BUG_ON(vcpu->kvm == NULL);
	kvm = vcpu->kvm;

3820
	vcpu->arch.mmu.root_hpa = INVALID_PAGE;
3821
	if (!irqchip_in_kernel(kvm) || vcpu->vcpu_id == 0)
3822
		vcpu->arch.mp_state = KVM_MP_STATE_RUNNABLE;
3823
	else
3824
		vcpu->arch.mp_state = KVM_MP_STATE_UNINITIALIZED;
3825 3826 3827 3828 3829 3830

	page = alloc_page(GFP_KERNEL | __GFP_ZERO);
	if (!page) {
		r = -ENOMEM;
		goto fail;
	}
3831
	vcpu->arch.pio_data = page_address(page);
3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847

	r = kvm_mmu_create(vcpu);
	if (r < 0)
		goto fail_free_pio_data;

	if (irqchip_in_kernel(kvm)) {
		r = kvm_create_lapic(vcpu);
		if (r < 0)
			goto fail_mmu_destroy;
	}

	return 0;

fail_mmu_destroy:
	kvm_mmu_destroy(vcpu);
fail_free_pio_data:
3848
	free_page((unsigned long)vcpu->arch.pio_data);
3849 3850 3851 3852 3853 3854 3855
fail:
	return r;
}

void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu)
{
	kvm_free_lapic(vcpu);
3856
	down_read(&vcpu->kvm->slots_lock);
3857
	kvm_mmu_destroy(vcpu);
3858
	up_read(&vcpu->kvm->slots_lock);
3859
	free_page((unsigned long)vcpu->arch.pio_data);
3860
}
3861 3862 3863 3864 3865 3866 3867 3868

struct  kvm *kvm_arch_create_vm(void)
{
	struct kvm *kvm = kzalloc(sizeof(struct kvm), GFP_KERNEL);

	if (!kvm)
		return ERR_PTR(-ENOMEM);

3869
	INIT_LIST_HEAD(&kvm->arch.active_mmu_pages);
3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901

	return kvm;
}

static void kvm_unload_vcpu_mmu(struct kvm_vcpu *vcpu)
{
	vcpu_load(vcpu);
	kvm_mmu_unload(vcpu);
	vcpu_put(vcpu);
}

static void kvm_free_vcpus(struct kvm *kvm)
{
	unsigned int i;

	/*
	 * Unpin any mmu pages first.
	 */
	for (i = 0; i < KVM_MAX_VCPUS; ++i)
		if (kvm->vcpus[i])
			kvm_unload_vcpu_mmu(kvm->vcpus[i]);
	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
		if (kvm->vcpus[i]) {
			kvm_arch_vcpu_free(kvm->vcpus[i]);
			kvm->vcpus[i] = NULL;
		}
	}

}

void kvm_arch_destroy_vm(struct kvm *kvm)
{
S
Sheng Yang 已提交
3902
	kvm_free_pit(kvm);
3903 3904
	kfree(kvm->arch.vpic);
	kfree(kvm->arch.vioapic);
3905 3906
	kvm_free_vcpus(kvm);
	kvm_free_physmem(kvm);
3907 3908
	if (kvm->arch.apic_access_page)
		put_page(kvm->arch.apic_access_page);
3909 3910
	kfree(kvm);
}
3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924

int kvm_arch_set_memory_region(struct kvm *kvm,
				struct kvm_userspace_memory_region *mem,
				struct kvm_memory_slot old,
				int user_alloc)
{
	int npages = mem->memory_size >> PAGE_SHIFT;
	struct kvm_memory_slot *memslot = &kvm->memslots[mem->slot];

	/*To keep backward compatibility with older userspace,
	 *x86 needs to hanlde !user_alloc case.
	 */
	if (!user_alloc) {
		if (npages && !old.rmap) {
3925
			down_write(&current->mm->mmap_sem);
3926 3927 3928 3929 3930
			memslot->userspace_addr = do_mmap(NULL, 0,
						     npages * PAGE_SIZE,
						     PROT_READ | PROT_WRITE,
						     MAP_SHARED | MAP_ANONYMOUS,
						     0);
3931
			up_write(&current->mm->mmap_sem);
3932 3933 3934 3935 3936 3937 3938

			if (IS_ERR((void *)memslot->userspace_addr))
				return PTR_ERR((void *)memslot->userspace_addr);
		} else {
			if (!old.user_alloc && old.rmap) {
				int ret;

3939
				down_write(&current->mm->mmap_sem);
3940 3941
				ret = do_munmap(current->mm, old.userspace_addr,
						old.npages * PAGE_SIZE);
3942
				up_write(&current->mm->mmap_sem);
3943 3944 3945 3946 3947 3948 3949 3950
				if (ret < 0)
					printk(KERN_WARNING
				       "kvm_vm_ioctl_set_memory_region: "
				       "failed to munmap memory\n");
			}
		}
	}

3951
	if (!kvm->arch.n_requested_mmu_pages) {
3952 3953 3954 3955 3956 3957 3958 3959 3960
		unsigned int nr_mmu_pages = kvm_mmu_calculate_mmu_pages(kvm);
		kvm_mmu_change_mmu_pages(kvm, nr_mmu_pages);
	}

	kvm_mmu_slot_remove_write_access(kvm, mem->slot);
	kvm_flush_remote_tlbs(kvm);

	return 0;
}
3961 3962 3963

int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
{
3964 3965
	return vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE
	       || vcpu->arch.mp_state == KVM_MP_STATE_SIPI_RECEIVED;
3966
}
3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978

static void vcpu_kick_intr(void *info)
{
#ifdef DEBUG
	struct kvm_vcpu *vcpu = (struct kvm_vcpu *)info;
	printk(KERN_DEBUG "vcpu_kick_intr %p \n", vcpu);
#endif
}

void kvm_vcpu_kick(struct kvm_vcpu *vcpu)
{
	int ipi_pcpu = vcpu->cpu;
3979
	int cpu = get_cpu();
3980 3981 3982 3983 3984

	if (waitqueue_active(&vcpu->wq)) {
		wake_up_interruptible(&vcpu->wq);
		++vcpu->stat.halt_wakeup;
	}
3985 3986 3987 3988 3989
	/*
	 * We may be called synchronously with irqs disabled in guest mode,
	 * So need not to call smp_call_function_single() in that case.
	 */
	if (vcpu->guest_mode && vcpu->cpu != cpu)
3990
		smp_call_function_single(ipi_pcpu, vcpu_kick_intr, vcpu, 0, 0);
3991
	put_cpu();
3992
}