paravirt.h 24.6 KB
Newer Older
H
H. Peter Anvin 已提交
1 2
#ifndef _ASM_X86_PARAVIRT_H
#define _ASM_X86_PARAVIRT_H
3 4
/* Various instructions on x86 need to be replaced for
 * para-virtualization: those hooks are defined here. */
5 6

#ifdef CONFIG_PARAVIRT
7
#include <asm/pgtable_types.h>
8
#include <asm/asm.h>
9

10
#include <asm/paravirt_types.h>
11

12
#ifndef __ASSEMBLY__
13
#include <linux/types.h>
14
#include <linux/cpumask.h>
15

16 17
static inline int paravirt_enabled(void)
{
18
	return pv_info.paravirt_enabled;
19
}
20

21
static inline void load_sp0(struct tss_struct *tss,
22 23
			     struct thread_struct *thread)
{
24
	PVOP_VCALL2(pv_cpu_ops.load_sp0, tss, thread);
25 26 27 28 29 30
}

/* The paravirtualized CPUID instruction. */
static inline void __cpuid(unsigned int *eax, unsigned int *ebx,
			   unsigned int *ecx, unsigned int *edx)
{
31
	PVOP_VCALL4(pv_cpu_ops.cpuid, eax, ebx, ecx, edx);
32 33 34 35 36
}

/*
 * These special macros can be used to get or set a debugging register
 */
37 38
static inline unsigned long paravirt_get_debugreg(int reg)
{
39
	return PVOP_CALL1(unsigned long, pv_cpu_ops.get_debugreg, reg);
40 41 42 43
}
#define get_debugreg(var, reg) var = paravirt_get_debugreg(reg)
static inline void set_debugreg(unsigned long val, int reg)
{
44
	PVOP_VCALL2(pv_cpu_ops.set_debugreg, reg, val);
45
}
46

47 48
static inline void clts(void)
{
49
	PVOP_VCALL0(pv_cpu_ops.clts);
50
}
51

52 53
static inline unsigned long read_cr0(void)
{
54
	return PVOP_CALL0(unsigned long, pv_cpu_ops.read_cr0);
55
}
56

57 58
static inline void write_cr0(unsigned long x)
{
59
	PVOP_VCALL1(pv_cpu_ops.write_cr0, x);
60 61 62 63
}

static inline unsigned long read_cr2(void)
{
64
	return PVOP_CALL0(unsigned long, pv_mmu_ops.read_cr2);
65 66 67 68
}

static inline void write_cr2(unsigned long x)
{
69
	PVOP_VCALL1(pv_mmu_ops.write_cr2, x);
70 71 72 73
}

static inline unsigned long read_cr3(void)
{
74
	return PVOP_CALL0(unsigned long, pv_mmu_ops.read_cr3);
75
}
76

77 78
static inline void write_cr3(unsigned long x)
{
79
	PVOP_VCALL1(pv_mmu_ops.write_cr3, x);
80
}
81

82 83
static inline unsigned long read_cr4(void)
{
84
	return PVOP_CALL0(unsigned long, pv_cpu_ops.read_cr4);
85 86 87
}
static inline unsigned long read_cr4_safe(void)
{
88
	return PVOP_CALL0(unsigned long, pv_cpu_ops.read_cr4_safe);
89
}
90

91 92
static inline void write_cr4(unsigned long x)
{
93
	PVOP_VCALL1(pv_cpu_ops.write_cr4, x);
94
}
95

96
#ifdef CONFIG_X86_64
97 98 99 100 101 102 103 104 105
static inline unsigned long read_cr8(void)
{
	return PVOP_CALL0(unsigned long, pv_cpu_ops.read_cr8);
}

static inline void write_cr8(unsigned long x)
{
	PVOP_VCALL1(pv_cpu_ops.write_cr8, x);
}
106
#endif
107

108 109
static inline void raw_safe_halt(void)
{
110
	PVOP_VCALL0(pv_irq_ops.safe_halt);
111 112 113 114
}

static inline void halt(void)
{
115
	PVOP_VCALL0(pv_irq_ops.safe_halt);
116 117 118 119
}

static inline void wbinvd(void)
{
120
	PVOP_VCALL0(pv_cpu_ops.wbinvd);
121 122
}

123
#define get_kernel_rpl()  (pv_info.kernel_rpl)
124

125 126
static inline u64 paravirt_read_msr(unsigned msr, int *err)
{
127
	return PVOP_CALL2(u64, pv_cpu_ops.read_msr, msr, err);
128
}
129 130 131 132 133 134

static inline int paravirt_rdmsr_regs(u32 *regs)
{
	return PVOP_CALL1(int, pv_cpu_ops.rdmsr_regs, regs);
}

135 136
static inline int paravirt_write_msr(unsigned msr, unsigned low, unsigned high)
{
137
	return PVOP_CALL3(int, pv_cpu_ops.write_msr, msr, low, high);
138 139
}

140 141 142 143 144
static inline int paravirt_wrmsr_regs(u32 *regs)
{
	return PVOP_CALL1(int, pv_cpu_ops.wrmsr_regs, regs);
}

145
/* These should all do BUG_ON(_err), but our headers are too tangled. */
146 147
#define rdmsr(msr, val1, val2)			\
do {						\
148 149 150 151
	int _err;				\
	u64 _l = paravirt_read_msr(msr, &_err);	\
	val1 = (u32)_l;				\
	val2 = _l >> 32;			\
152
} while (0)
153

154 155
#define wrmsr(msr, val1, val2)			\
do {						\
156
	paravirt_write_msr(msr, val1, val2);	\
157
} while (0)
158

159 160
#define rdmsrl(msr, val)			\
do {						\
161 162
	int _err;				\
	val = paravirt_read_msr(msr, &_err);	\
163
} while (0)
164

165 166
#define wrmsrl(msr, val)	wrmsr(msr, (u32)((u64)(val)), ((u64)(val))>>32)
#define wrmsr_safe(msr, a, b)	paravirt_write_msr(msr, a, b)
167 168

/* rdmsr with exception handling */
169 170
#define rdmsr_safe(msr, a, b)			\
({						\
171 172 173 174
	int _err;				\
	u64 _l = paravirt_read_msr(msr, &_err);	\
	(*a) = (u32)_l;				\
	(*b) = _l >> 32;			\
175 176
	_err;					\
})
177

178 179 180
#define rdmsr_safe_regs(regs)	paravirt_rdmsr_regs(regs)
#define wrmsr_safe_regs(regs)	paravirt_wrmsr_regs(regs)

A
Andi Kleen 已提交
181 182 183 184 185 186 187
static inline int rdmsrl_safe(unsigned msr, unsigned long long *p)
{
	int err;

	*p = paravirt_read_msr(msr, &err);
	return err;
}
Y
Yinghai Lu 已提交
188 189
static inline int rdmsrl_amd_safe(unsigned msr, unsigned long long *p)
{
190
	u32 gprs[8] = { 0 };
Y
Yinghai Lu 已提交
191 192
	int err;

193 194 195 196 197 198 199
	gprs[1] = msr;
	gprs[7] = 0x9c5a203a;

	err = paravirt_rdmsr_regs(gprs);

	*p = gprs[0] | ((u64)gprs[2] << 32);

Y
Yinghai Lu 已提交
200 201
	return err;
}
202

203 204 205 206 207 208 209 210 211 212 213 214
static inline int wrmsrl_amd_safe(unsigned msr, unsigned long long val)
{
	u32 gprs[8] = { 0 };

	gprs[0] = (u32)val;
	gprs[1] = msr;
	gprs[2] = val >> 32;
	gprs[7] = 0x9c5a203a;

	return paravirt_wrmsr_regs(gprs);
}

215 216
static inline u64 paravirt_read_tsc(void)
{
217
	return PVOP_CALL0(u64, pv_cpu_ops.read_tsc);
218
}
219

220 221
#define rdtscl(low)				\
do {						\
222 223
	u64 _l = paravirt_read_tsc();		\
	low = (int)_l;				\
224
} while (0)
225

226
#define rdtscll(val) (val = paravirt_read_tsc())
227

228 229
static inline unsigned long long paravirt_sched_clock(void)
{
230
	return PVOP_CALL0(unsigned long long, pv_time_ops.sched_clock);
231
}
232

233 234
static inline unsigned long long paravirt_read_pmc(int counter)
{
235
	return PVOP_CALL1(u64, pv_cpu_ops.read_pmc, counter);
236
}
237

238 239
#define rdpmc(counter, low, high)		\
do {						\
240 241 242
	u64 _l = paravirt_read_pmc(counter);	\
	low = (u32)_l;				\
	high = _l >> 32;			\
243
} while (0)
244

245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
static inline unsigned long long paravirt_rdtscp(unsigned int *aux)
{
	return PVOP_CALL1(u64, pv_cpu_ops.read_tscp, aux);
}

#define rdtscp(low, high, aux)				\
do {							\
	int __aux;					\
	unsigned long __val = paravirt_rdtscp(&__aux);	\
	(low) = (u32)__val;				\
	(high) = (u32)(__val >> 32);			\
	(aux) = __aux;					\
} while (0)

#define rdtscpll(val, aux)				\
do {							\
	unsigned long __aux; 				\
	val = paravirt_rdtscp(&__aux);			\
	(aux) = __aux;					\
} while (0)

266 267 268 269 270 271 272 273 274 275
static inline void paravirt_alloc_ldt(struct desc_struct *ldt, unsigned entries)
{
	PVOP_VCALL2(pv_cpu_ops.alloc_ldt, ldt, entries);
}

static inline void paravirt_free_ldt(struct desc_struct *ldt, unsigned entries)
{
	PVOP_VCALL2(pv_cpu_ops.free_ldt, ldt, entries);
}

276 277
static inline void load_TR_desc(void)
{
278
	PVOP_VCALL0(pv_cpu_ops.load_tr_desc);
279
}
280
static inline void load_gdt(const struct desc_ptr *dtr)
281
{
282
	PVOP_VCALL1(pv_cpu_ops.load_gdt, dtr);
283
}
284
static inline void load_idt(const struct desc_ptr *dtr)
285
{
286
	PVOP_VCALL1(pv_cpu_ops.load_idt, dtr);
287 288 289
}
static inline void set_ldt(const void *addr, unsigned entries)
{
290
	PVOP_VCALL2(pv_cpu_ops.set_ldt, addr, entries);
291
}
292
static inline void store_gdt(struct desc_ptr *dtr)
293
{
294
	PVOP_VCALL1(pv_cpu_ops.store_gdt, dtr);
295
}
296
static inline void store_idt(struct desc_ptr *dtr)
297
{
298
	PVOP_VCALL1(pv_cpu_ops.store_idt, dtr);
299 300 301
}
static inline unsigned long paravirt_store_tr(void)
{
302
	return PVOP_CALL0(unsigned long, pv_cpu_ops.store_tr);
303 304 305 306
}
#define store_tr(tr)	((tr) = paravirt_store_tr())
static inline void load_TLS(struct thread_struct *t, unsigned cpu)
{
307
	PVOP_VCALL2(pv_cpu_ops.load_tls, t, cpu);
308
}
309

310 311 312 313 314 315 316
#ifdef CONFIG_X86_64
static inline void load_gs_index(unsigned int gs)
{
	PVOP_VCALL1(pv_cpu_ops.load_gs_index, gs);
}
#endif

317 318
static inline void write_ldt_entry(struct desc_struct *dt, int entry,
				   const void *desc)
319
{
320
	PVOP_VCALL3(pv_cpu_ops.write_ldt_entry, dt, entry, desc);
321
}
322 323 324

static inline void write_gdt_entry(struct desc_struct *dt, int entry,
				   void *desc, int type)
325
{
326
	PVOP_VCALL4(pv_cpu_ops.write_gdt_entry, dt, entry, desc, type);
327
}
328

329
static inline void write_idt_entry(gate_desc *dt, int entry, const gate_desc *g)
330
{
331
	PVOP_VCALL3(pv_cpu_ops.write_idt_entry, dt, entry, g);
332 333 334
}
static inline void set_iopl_mask(unsigned mask)
{
335
	PVOP_VCALL1(pv_cpu_ops.set_iopl_mask, mask);
336
}
337

338
/* The paravirtualized I/O functions */
339 340
static inline void slow_down_io(void)
{
341
	pv_cpu_ops.io_delay();
342
#ifdef REALLY_SLOW_IO
343 344 345
	pv_cpu_ops.io_delay();
	pv_cpu_ops.io_delay();
	pv_cpu_ops.io_delay();
346 347 348
#endif
}

349 350 351 352
#ifdef CONFIG_SMP
static inline void startup_ipi_hook(int phys_apicid, unsigned long start_eip,
				    unsigned long start_esp)
{
353 354
	PVOP_VCALL3(pv_apic_ops.startup_ipi_hook,
		    phys_apicid, start_eip, start_esp);
355 356
}
#endif
357

358 359 360
static inline void paravirt_activate_mm(struct mm_struct *prev,
					struct mm_struct *next)
{
361
	PVOP_VCALL2(pv_mmu_ops.activate_mm, prev, next);
362 363 364 365 366
}

static inline void arch_dup_mmap(struct mm_struct *oldmm,
				 struct mm_struct *mm)
{
367
	PVOP_VCALL2(pv_mmu_ops.dup_mmap, oldmm, mm);
368 369 370 371
}

static inline void arch_exit_mmap(struct mm_struct *mm)
{
372
	PVOP_VCALL1(pv_mmu_ops.exit_mmap, mm);
373 374
}

375 376
static inline void __flush_tlb(void)
{
377
	PVOP_VCALL0(pv_mmu_ops.flush_tlb_user);
378 379 380
}
static inline void __flush_tlb_global(void)
{
381
	PVOP_VCALL0(pv_mmu_ops.flush_tlb_kernel);
382 383 384
}
static inline void __flush_tlb_single(unsigned long addr)
{
385
	PVOP_VCALL1(pv_mmu_ops.flush_tlb_single, addr);
386
}
387

388 389
static inline void flush_tlb_others(const struct cpumask *cpumask,
				    struct mm_struct *mm,
390 391
				    unsigned long va)
{
392
	PVOP_VCALL3(pv_mmu_ops.flush_tlb_others, cpumask, mm, va);
393 394
}

395 396 397 398 399 400 401 402 403 404
static inline int paravirt_pgd_alloc(struct mm_struct *mm)
{
	return PVOP_CALL1(int, pv_mmu_ops.pgd_alloc, mm);
}

static inline void paravirt_pgd_free(struct mm_struct *mm, pgd_t *pgd)
{
	PVOP_VCALL2(pv_mmu_ops.pgd_free, mm, pgd);
}

405
static inline void paravirt_alloc_pte(struct mm_struct *mm, unsigned long pfn)
406
{
407
	PVOP_VCALL2(pv_mmu_ops.alloc_pte, mm, pfn);
408
}
409
static inline void paravirt_release_pte(unsigned long pfn)
410
{
411
	PVOP_VCALL1(pv_mmu_ops.release_pte, pfn);
412
}
413

414
static inline void paravirt_alloc_pmd(struct mm_struct *mm, unsigned long pfn)
415
{
416
	PVOP_VCALL2(pv_mmu_ops.alloc_pmd, mm, pfn);
417
}
418

419 420
static inline void paravirt_alloc_pmd_clone(unsigned long pfn, unsigned long clonepfn,
					    unsigned long start, unsigned long count)
421
{
422
	PVOP_VCALL4(pv_mmu_ops.alloc_pmd_clone, pfn, clonepfn, start, count);
423
}
424
static inline void paravirt_release_pmd(unsigned long pfn)
425
{
426
	PVOP_VCALL1(pv_mmu_ops.release_pmd, pfn);
427 428
}

429
static inline void paravirt_alloc_pud(struct mm_struct *mm, unsigned long pfn)
430 431 432
{
	PVOP_VCALL2(pv_mmu_ops.alloc_pud, mm, pfn);
}
433
static inline void paravirt_release_pud(unsigned long pfn)
434 435 436 437
{
	PVOP_VCALL1(pv_mmu_ops.release_pud, pfn);
}

438 439
static inline void pte_update(struct mm_struct *mm, unsigned long addr,
			      pte_t *ptep)
440
{
441
	PVOP_VCALL3(pv_mmu_ops.pte_update, mm, addr, ptep);
442 443
}

444 445
static inline void pte_update_defer(struct mm_struct *mm, unsigned long addr,
				    pte_t *ptep)
446
{
447
	PVOP_VCALL3(pv_mmu_ops.pte_update_defer, mm, addr, ptep);
448 449
}

450
static inline pte_t __pte(pteval_t val)
451
{
452 453 454
	pteval_t ret;

	if (sizeof(pteval_t) > sizeof(long))
455 456 457
		ret = PVOP_CALLEE2(pteval_t,
				   pv_mmu_ops.make_pte,
				   val, (u64)val >> 32);
458
	else
459 460 461
		ret = PVOP_CALLEE1(pteval_t,
				   pv_mmu_ops.make_pte,
				   val);
462

463
	return (pte_t) { .pte = ret };
464 465
}

466 467 468 469 470
static inline pteval_t pte_val(pte_t pte)
{
	pteval_t ret;

	if (sizeof(pteval_t) > sizeof(long))
471 472
		ret = PVOP_CALLEE2(pteval_t, pv_mmu_ops.pte_val,
				   pte.pte, (u64)pte.pte >> 32);
473
	else
474 475
		ret = PVOP_CALLEE1(pteval_t, pv_mmu_ops.pte_val,
				   pte.pte);
476 477 478 479

	return ret;
}

480
static inline pgd_t __pgd(pgdval_t val)
481
{
482 483 484
	pgdval_t ret;

	if (sizeof(pgdval_t) > sizeof(long))
485 486
		ret = PVOP_CALLEE2(pgdval_t, pv_mmu_ops.make_pgd,
				   val, (u64)val >> 32);
487
	else
488 489
		ret = PVOP_CALLEE1(pgdval_t, pv_mmu_ops.make_pgd,
				   val);
490 491 492 493 494 495 496 497 498

	return (pgd_t) { ret };
}

static inline pgdval_t pgd_val(pgd_t pgd)
{
	pgdval_t ret;

	if (sizeof(pgdval_t) > sizeof(long))
499 500
		ret =  PVOP_CALLEE2(pgdval_t, pv_mmu_ops.pgd_val,
				    pgd.pgd, (u64)pgd.pgd >> 32);
501
	else
502 503
		ret =  PVOP_CALLEE1(pgdval_t, pv_mmu_ops.pgd_val,
				    pgd.pgd);
504 505

	return ret;
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
#define  __HAVE_ARCH_PTEP_MODIFY_PROT_TRANSACTION
static inline pte_t ptep_modify_prot_start(struct mm_struct *mm, unsigned long addr,
					   pte_t *ptep)
{
	pteval_t ret;

	ret = PVOP_CALL3(pteval_t, pv_mmu_ops.ptep_modify_prot_start,
			 mm, addr, ptep);

	return (pte_t) { .pte = ret };
}

static inline void ptep_modify_prot_commit(struct mm_struct *mm, unsigned long addr,
					   pte_t *ptep, pte_t pte)
{
	if (sizeof(pteval_t) > sizeof(long))
		/* 5 arg words */
		pv_mmu_ops.ptep_modify_prot_commit(mm, addr, ptep, pte);
	else
		PVOP_VCALL4(pv_mmu_ops.ptep_modify_prot_commit,
			    mm, addr, ptep, pte.pte);
}

531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
static inline void set_pte(pte_t *ptep, pte_t pte)
{
	if (sizeof(pteval_t) > sizeof(long))
		PVOP_VCALL3(pv_mmu_ops.set_pte, ptep,
			    pte.pte, (u64)pte.pte >> 32);
	else
		PVOP_VCALL2(pv_mmu_ops.set_pte, ptep,
			    pte.pte);
}

static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
			      pte_t *ptep, pte_t pte)
{
	if (sizeof(pteval_t) > sizeof(long))
		/* 5 arg words */
		pv_mmu_ops.set_pte_at(mm, addr, ptep, pte);
	else
		PVOP_VCALL4(pv_mmu_ops.set_pte_at, mm, addr, ptep, pte.pte);
}

551 552 553 554 555 556 557 558 559 560
static inline void set_pmd(pmd_t *pmdp, pmd_t pmd)
{
	pmdval_t val = native_pmd_val(pmd);

	if (sizeof(pmdval_t) > sizeof(long))
		PVOP_VCALL3(pv_mmu_ops.set_pmd, pmdp, val, (u64)val >> 32);
	else
		PVOP_VCALL2(pv_mmu_ops.set_pmd, pmdp, val);
}

561 562 563 564 565 566
#if PAGETABLE_LEVELS >= 3
static inline pmd_t __pmd(pmdval_t val)
{
	pmdval_t ret;

	if (sizeof(pmdval_t) > sizeof(long))
567 568
		ret = PVOP_CALLEE2(pmdval_t, pv_mmu_ops.make_pmd,
				   val, (u64)val >> 32);
569
	else
570 571
		ret = PVOP_CALLEE1(pmdval_t, pv_mmu_ops.make_pmd,
				   val);
572 573 574 575 576 577 578 579 580

	return (pmd_t) { ret };
}

static inline pmdval_t pmd_val(pmd_t pmd)
{
	pmdval_t ret;

	if (sizeof(pmdval_t) > sizeof(long))
581 582
		ret =  PVOP_CALLEE2(pmdval_t, pv_mmu_ops.pmd_val,
				    pmd.pmd, (u64)pmd.pmd >> 32);
583
	else
584 585
		ret =  PVOP_CALLEE1(pmdval_t, pv_mmu_ops.pmd_val,
				    pmd.pmd);
586 587 588 589 590 591 592 593 594 595 596 597 598 599 600

	return ret;
}

static inline void set_pud(pud_t *pudp, pud_t pud)
{
	pudval_t val = native_pud_val(pud);

	if (sizeof(pudval_t) > sizeof(long))
		PVOP_VCALL3(pv_mmu_ops.set_pud, pudp,
			    val, (u64)val >> 32);
	else
		PVOP_VCALL2(pv_mmu_ops.set_pud, pudp,
			    val);
}
601 602 603 604 605 606
#if PAGETABLE_LEVELS == 4
static inline pud_t __pud(pudval_t val)
{
	pudval_t ret;

	if (sizeof(pudval_t) > sizeof(long))
607 608
		ret = PVOP_CALLEE2(pudval_t, pv_mmu_ops.make_pud,
				   val, (u64)val >> 32);
609
	else
610 611
		ret = PVOP_CALLEE1(pudval_t, pv_mmu_ops.make_pud,
				   val);
612 613 614 615 616 617 618 619 620

	return (pud_t) { ret };
}

static inline pudval_t pud_val(pud_t pud)
{
	pudval_t ret;

	if (sizeof(pudval_t) > sizeof(long))
621 622
		ret =  PVOP_CALLEE2(pudval_t, pv_mmu_ops.pud_val,
				    pud.pud, (u64)pud.pud >> 32);
623
	else
624 625
		ret =  PVOP_CALLEE1(pudval_t, pv_mmu_ops.pud_val,
				    pud.pud);
626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653

	return ret;
}

static inline void set_pgd(pgd_t *pgdp, pgd_t pgd)
{
	pgdval_t val = native_pgd_val(pgd);

	if (sizeof(pgdval_t) > sizeof(long))
		PVOP_VCALL3(pv_mmu_ops.set_pgd, pgdp,
			    val, (u64)val >> 32);
	else
		PVOP_VCALL2(pv_mmu_ops.set_pgd, pgdp,
			    val);
}

static inline void pgd_clear(pgd_t *pgdp)
{
	set_pgd(pgdp, __pgd(0));
}

static inline void pud_clear(pud_t *pudp)
{
	set_pud(pudp, __pud(0));
}

#endif	/* PAGETABLE_LEVELS == 4 */

654 655
#endif	/* PAGETABLE_LEVELS >= 3 */

656 657 658 659 660 661 662 663 664 665 666 667 668 669
#ifdef CONFIG_X86_PAE
/* Special-case pte-setting operations for PAE, which can't update a
   64-bit pte atomically */
static inline void set_pte_atomic(pte_t *ptep, pte_t pte)
{
	PVOP_VCALL3(pv_mmu_ops.set_pte_atomic, ptep,
		    pte.pte, pte.pte >> 32);
}

static inline void pte_clear(struct mm_struct *mm, unsigned long addr,
			     pte_t *ptep)
{
	PVOP_VCALL3(pv_mmu_ops.pte_clear, mm, addr, ptep);
}
670 671 672 673 674

static inline void pmd_clear(pmd_t *pmdp)
{
	PVOP_VCALL1(pv_mmu_ops.pmd_clear, pmdp);
}
675 676 677 678 679 680 681 682 683 684 685
#else  /* !CONFIG_X86_PAE */
static inline void set_pte_atomic(pte_t *ptep, pte_t pte)
{
	set_pte(ptep, pte);
}

static inline void pte_clear(struct mm_struct *mm, unsigned long addr,
			     pte_t *ptep)
{
	set_pte_at(mm, addr, ptep, __pte(0));
}
686 687 688 689 690

static inline void pmd_clear(pmd_t *pmdp)
{
	set_pmd(pmdp, __pmd(0));
}
691 692
#endif	/* CONFIG_X86_PAE */

693
#define  __HAVE_ARCH_START_CONTEXT_SWITCH
694
static inline void arch_start_context_switch(struct task_struct *prev)
695
{
696
	PVOP_VCALL1(pv_cpu_ops.start_context_switch, prev);
697 698
}

699
static inline void arch_end_context_switch(struct task_struct *next)
700
{
701
	PVOP_VCALL1(pv_cpu_ops.end_context_switch, next);
702 703
}

704
#define  __HAVE_ARCH_ENTER_LAZY_MMU_MODE
705 706
static inline void arch_enter_lazy_mmu_mode(void)
{
707
	PVOP_VCALL0(pv_mmu_ops.lazy_mode.enter);
708 709 710 711
}

static inline void arch_leave_lazy_mmu_mode(void)
{
712
	PVOP_VCALL0(pv_mmu_ops.lazy_mode.leave);
713 714
}

715
void arch_flush_lazy_mmu_mode(void);
716

717
static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx,
718
				phys_addr_t phys, pgprot_t flags)
719 720 721 722
{
	pv_mmu_ops.set_fixmap(idx, phys, flags);
}

723
#if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS)
724

725
static inline int arch_spin_is_locked(struct arch_spinlock *lock)
726 727 728 729
{
	return PVOP_CALL1(int, pv_lock_ops.spin_is_locked, lock);
}

730
static inline int arch_spin_is_contended(struct arch_spinlock *lock)
731 732 733
{
	return PVOP_CALL1(int, pv_lock_ops.spin_is_contended, lock);
}
734
#define arch_spin_is_contended	arch_spin_is_contended
735

736
static __always_inline void arch_spin_lock(struct arch_spinlock *lock)
737
{
738
	PVOP_VCALL1(pv_lock_ops.spin_lock, lock);
739 740
}

741
static __always_inline void arch_spin_lock_flags(struct arch_spinlock *lock,
742 743 744 745 746
						  unsigned long flags)
{
	PVOP_VCALL2(pv_lock_ops.spin_lock_flags, lock, flags);
}

747
static __always_inline int arch_spin_trylock(struct arch_spinlock *lock)
748 749 750 751
{
	return PVOP_CALL1(int, pv_lock_ops.spin_trylock, lock);
}

752
static __always_inline void arch_spin_unlock(struct arch_spinlock *lock)
753
{
754
	PVOP_VCALL1(pv_lock_ops.spin_unlock, lock);
755 756
}

757 758
#endif

759
#ifdef CONFIG_X86_32
760 761 762 763
#define PV_SAVE_REGS "pushl %ecx; pushl %edx;"
#define PV_RESTORE_REGS "popl %edx; popl %ecx;"

/* save and restore all caller-save registers, except return value */
764 765
#define PV_SAVE_ALL_CALLER_REGS		"pushl %ecx;"
#define PV_RESTORE_ALL_CALLER_REGS	"popl  %ecx;"
766

767 768 769 770
#define PV_FLAGS_ARG "0"
#define PV_EXTRA_CLOBBERS
#define PV_VEXTRA_CLOBBERS
#else
771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790
/* save and restore all caller-save registers, except return value */
#define PV_SAVE_ALL_CALLER_REGS						\
	"push %rcx;"							\
	"push %rdx;"							\
	"push %rsi;"							\
	"push %rdi;"							\
	"push %r8;"							\
	"push %r9;"							\
	"push %r10;"							\
	"push %r11;"
#define PV_RESTORE_ALL_CALLER_REGS					\
	"pop %r11;"							\
	"pop %r10;"							\
	"pop %r9;"							\
	"pop %r8;"							\
	"pop %rdi;"							\
	"pop %rsi;"							\
	"pop %rdx;"							\
	"pop %rcx;"

791 792 793 794
/* We save some registers, but all of them, that's too much. We clobber all
 * caller saved registers but the argument parameter */
#define PV_SAVE_REGS "pushq %%rdi;"
#define PV_RESTORE_REGS "popq %%rdi;"
795 796
#define PV_EXTRA_CLOBBERS EXTRA_CLOBBERS, "rcx" , "rdx", "rsi"
#define PV_VEXTRA_CLOBBERS EXTRA_CLOBBERS, "rdi", "rcx" , "rdx", "rsi"
797 798 799
#define PV_FLAGS_ARG "D"
#endif

800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831
/*
 * Generate a thunk around a function which saves all caller-save
 * registers except for the return value.  This allows C functions to
 * be called from assembler code where fewer than normal registers are
 * available.  It may also help code generation around calls from C
 * code if the common case doesn't use many registers.
 *
 * When a callee is wrapped in a thunk, the caller can assume that all
 * arg regs and all scratch registers are preserved across the
 * call. The return value in rax/eax will not be saved, even for void
 * functions.
 */
#define PV_CALLEE_SAVE_REGS_THUNK(func)					\
	extern typeof(func) __raw_callee_save_##func;			\
	static void *__##func##__ __used = func;			\
									\
	asm(".pushsection .text;"					\
	    "__raw_callee_save_" #func ": "				\
	    PV_SAVE_ALL_CALLER_REGS					\
	    "call " #func ";"						\
	    PV_RESTORE_ALL_CALLER_REGS					\
	    "ret;"							\
	    ".popsection")

/* Get a reference to a callee-save function */
#define PV_CALLEE_SAVE(func)						\
	((struct paravirt_callee_save) { __raw_callee_save_##func })

/* Promise that "func" already uses the right calling convention */
#define __PV_IS_CALLEE_SAVE(func)			\
	((struct paravirt_callee_save) { func })

832 833
static inline unsigned long __raw_local_save_flags(void)
{
834
	return PVOP_CALLEE0(unsigned long, pv_irq_ops.save_fl);
835 836 837 838
}

static inline void raw_local_irq_restore(unsigned long f)
{
839
	PVOP_VCALLEE1(pv_irq_ops.restore_fl, f);
840 841 842 843
}

static inline void raw_local_irq_disable(void)
{
844
	PVOP_VCALLEE0(pv_irq_ops.irq_disable);
845 846 847 848
}

static inline void raw_local_irq_enable(void)
{
849
	PVOP_VCALLEE0(pv_irq_ops.irq_enable);
850 851 852 853 854 855
}

static inline unsigned long __raw_local_irq_save(void)
{
	unsigned long f;

856 857
	f = __raw_local_save_flags();
	raw_local_irq_disable();
858 859 860
	return f;
}

861

862
/* Make sure as little as possible of this mess escapes. */
863
#undef PARAVIRT_CALL
864 865
#undef __PVOP_CALL
#undef __PVOP_VCALL
866 867 868 869 870 871 872 873 874 875
#undef PVOP_VCALL0
#undef PVOP_CALL0
#undef PVOP_VCALL1
#undef PVOP_CALL1
#undef PVOP_VCALL2
#undef PVOP_CALL2
#undef PVOP_VCALL3
#undef PVOP_CALL3
#undef PVOP_VCALL4
#undef PVOP_CALL4
876

877 878
extern void default_banner(void);

879 880
#else  /* __ASSEMBLY__ */

881
#define _PVSITE(ptype, clobbers, ops, word, algn)	\
882 883 884 885
771:;						\
	ops;					\
772:;						\
	.pushsection .parainstructions,"a";	\
886 887
	 .align	algn;				\
	 word 771b;				\
888 889 890 891 892
	 .byte ptype;				\
	 .byte 772b-771b;			\
	 .short clobbers;			\
	.popsection

893

894
#define COND_PUSH(set, mask, reg)			\
895
	.if ((~(set)) & mask); push %reg; .endif
896
#define COND_POP(set, mask, reg)			\
897
	.if ((~(set)) & mask); pop %reg; .endif
898

899
#ifdef CONFIG_X86_64
900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921

#define PV_SAVE_REGS(set)			\
	COND_PUSH(set, CLBR_RAX, rax);		\
	COND_PUSH(set, CLBR_RCX, rcx);		\
	COND_PUSH(set, CLBR_RDX, rdx);		\
	COND_PUSH(set, CLBR_RSI, rsi);		\
	COND_PUSH(set, CLBR_RDI, rdi);		\
	COND_PUSH(set, CLBR_R8, r8);		\
	COND_PUSH(set, CLBR_R9, r9);		\
	COND_PUSH(set, CLBR_R10, r10);		\
	COND_PUSH(set, CLBR_R11, r11)
#define PV_RESTORE_REGS(set)			\
	COND_POP(set, CLBR_R11, r11);		\
	COND_POP(set, CLBR_R10, r10);		\
	COND_POP(set, CLBR_R9, r9);		\
	COND_POP(set, CLBR_R8, r8);		\
	COND_POP(set, CLBR_RDI, rdi);		\
	COND_POP(set, CLBR_RSI, rsi);		\
	COND_POP(set, CLBR_RDX, rdx);		\
	COND_POP(set, CLBR_RCX, rcx);		\
	COND_POP(set, CLBR_RAX, rax)

922
#define PARA_PATCH(struct, off)        ((PARAVIRT_PATCH_##struct + (off)) / 8)
923
#define PARA_SITE(ptype, clobbers, ops) _PVSITE(ptype, clobbers, ops, .quad, 8)
924
#define PARA_INDIRECT(addr)	*addr(%rip)
925
#else
926 927 928 929 930 931 932 933 934 935 936
#define PV_SAVE_REGS(set)			\
	COND_PUSH(set, CLBR_EAX, eax);		\
	COND_PUSH(set, CLBR_EDI, edi);		\
	COND_PUSH(set, CLBR_ECX, ecx);		\
	COND_PUSH(set, CLBR_EDX, edx)
#define PV_RESTORE_REGS(set)			\
	COND_POP(set, CLBR_EDX, edx);		\
	COND_POP(set, CLBR_ECX, ecx);		\
	COND_POP(set, CLBR_EDI, edi);		\
	COND_POP(set, CLBR_EAX, eax)

937
#define PARA_PATCH(struct, off)        ((PARAVIRT_PATCH_##struct + (off)) / 4)
938
#define PARA_SITE(ptype, clobbers, ops) _PVSITE(ptype, clobbers, ops, .long, 4)
939
#define PARA_INDIRECT(addr)	*%cs:addr
940 941
#endif

942 943
#define INTERRUPT_RETURN						\
	PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_iret), CLBR_NONE,	\
944
		  jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_iret))
945 946

#define DISABLE_INTERRUPTS(clobbers)					\
947
	PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_disable), clobbers, \
948
		  PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE);		\
949
		  call PARA_INDIRECT(pv_irq_ops+PV_IRQ_irq_disable);	\
950
		  PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);)
951 952

#define ENABLE_INTERRUPTS(clobbers)					\
953
	PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_enable), clobbers,	\
954
		  PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE);		\
955
		  call PARA_INDIRECT(pv_irq_ops+PV_IRQ_irq_enable);	\
956
		  PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);)
957

958 959
#define USERGS_SYSRET32							\
	PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_usergs_sysret32),	\
960
		  CLBR_NONE,						\
961
		  jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_usergs_sysret32))
962

963
#ifdef CONFIG_X86_32
964 965 966
#define GET_CR0_INTO_EAX				\
	push %ecx; push %edx;				\
	call PARA_INDIRECT(pv_cpu_ops+PV_CPU_read_cr0);	\
967
	pop %edx; pop %ecx
968 969 970 971 972 973 974 975

#define ENABLE_INTERRUPTS_SYSEXIT					\
	PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_irq_enable_sysexit),	\
		  CLBR_NONE,						\
		  jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_irq_enable_sysexit))


#else	/* !CONFIG_X86_32 */
976 977 978 979 980 981 982 983 984 985

/*
 * If swapgs is used while the userspace stack is still current,
 * there's no way to call a pvop.  The PV replacement *must* be
 * inlined, or the swapgs instruction must be trapped and emulated.
 */
#define SWAPGS_UNSAFE_STACK						\
	PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), CLBR_NONE,	\
		  swapgs)

986 987 988 989 990 991
/*
 * Note: swapgs is very special, and in practise is either going to be
 * implemented with a single "swapgs" instruction or something very
 * special.  Either way, we don't need to save any registers for
 * it.
 */
992 993
#define SWAPGS								\
	PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), CLBR_NONE,	\
994
		  call PARA_INDIRECT(pv_cpu_ops+PV_CPU_swapgs)		\
995 996
		 )

997 998 999
#define GET_CR2_INTO_RCX				\
	call PARA_INDIRECT(pv_mmu_ops+PV_MMU_read_cr2);	\
	movq %rax, %rcx;				\
1000 1001
	xorq %rax, %rax;

1002 1003 1004 1005 1006
#define PARAVIRT_ADJUST_EXCEPTION_FRAME					\
	PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_adjust_exception_frame), \
		  CLBR_NONE,						\
		  call PARA_INDIRECT(pv_irq_ops+PV_IRQ_adjust_exception_frame))

1007 1008
#define USERGS_SYSRET64							\
	PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_usergs_sysret64),	\
1009
		  CLBR_NONE,						\
1010 1011 1012 1013 1014 1015 1016
		  jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_usergs_sysret64))

#define ENABLE_INTERRUPTS_SYSEXIT32					\
	PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_irq_enable_sysexit),	\
		  CLBR_NONE,						\
		  jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_irq_enable_sysexit))
#endif	/* CONFIG_X86_32 */
1017

1018
#endif /* __ASSEMBLY__ */
1019 1020 1021
#else  /* CONFIG_PARAVIRT */
# define default_banner x86_init_noop
#endif /* !CONFIG_PARAVIRT */
H
H. Peter Anvin 已提交
1022
#endif /* _ASM_X86_PARAVIRT_H */