paravirt.c 11.7 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
/*  Paravirtualization interfaces
    Copyright (C) 2006 Rusty Russell IBM Corporation

    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation; either version 2 of the License, or
    (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program; if not, write to the Free Software
    Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
17 18

    2007 - x86_64 support added by Glauber de Oliveira Costa, Red Hat Inc
19
*/
20

21 22 23 24
#include <linux/errno.h>
#include <linux/module.h>
#include <linux/efi.h>
#include <linux/bcd.h>
25
#include <linux/highmem.h>
26 27 28 29 30 31 32 33 34

#include <asm/bug.h>
#include <asm/paravirt.h>
#include <asm/desc.h>
#include <asm/setup.h>
#include <asm/arch_hooks.h>
#include <asm/time.h>
#include <asm/irq.h>
#include <asm/delay.h>
35 36
#include <asm/fixmap.h>
#include <asm/apic.h>
37
#include <asm/tlbflush.h>
38
#include <asm/timer.h>
39 40

/* nop stub */
41
void _paravirt_nop(void)
42 43 44 45 46 47
{
}

static void __init default_banner(void)
{
	printk(KERN_INFO "Booting paravirtualized kernel on %s\n",
48
	       pv_info.name);
49 50 51 52
}

char *memory_setup(void)
{
53
	return pv_init_ops.memory_setup();
54 55
}

56
/* Simple instruction patching code. */
57 58 59 60 61 62 63 64 65
#define DEF_NATIVE(ops, name, code)					\
	extern const char start_##ops##_##name[], end_##ops##_##name[];	\
	asm("start_" #ops "_" #name ": " code "; end_" #ops "_" #name ":")

DEF_NATIVE(pv_irq_ops, irq_disable, "cli");
DEF_NATIVE(pv_irq_ops, irq_enable, "sti");
DEF_NATIVE(pv_irq_ops, restore_fl, "push %eax; popf");
DEF_NATIVE(pv_irq_ops, save_fl, "pushf; pop %eax");
DEF_NATIVE(pv_cpu_ops, iret, "iret");
66
DEF_NATIVE(pv_cpu_ops, irq_enable_syscall_ret, "sti; sysexit");
67 68 69 70 71 72 73 74
DEF_NATIVE(pv_mmu_ops, read_cr2, "mov %cr2, %eax");
DEF_NATIVE(pv_mmu_ops, write_cr3, "mov %eax, %cr3");
DEF_NATIVE(pv_mmu_ops, read_cr3, "mov %cr3, %eax");
DEF_NATIVE(pv_cpu_ops, clts, "clts");
DEF_NATIVE(pv_cpu_ops, read_tsc, "rdtsc");

/* Undefined instruction for dealing with missing ops pointers. */
static const unsigned char ud2a[] = { 0x0f, 0x0b };
75

76 77
static unsigned native_patch(u8 type, u16 clobbers, void *ibuf,
			     unsigned long addr, unsigned len)
78
{
79 80 81 82
	const unsigned char *start, *end;
	unsigned ret;

	switch(type) {
83 84 85 86 87 88 89 90 91 92 93
#define SITE(ops, x)						\
	case PARAVIRT_PATCH(ops.x):				\
		start = start_##ops##_##x;			\
		end = end_##ops##_##x;				\
		goto patch_site

	SITE(pv_irq_ops, irq_disable);
	SITE(pv_irq_ops, irq_enable);
	SITE(pv_irq_ops, restore_fl);
	SITE(pv_irq_ops, save_fl);
	SITE(pv_cpu_ops, iret);
94
	SITE(pv_cpu_ops, irq_enable_syscall_ret);
95 96 97 98 99
	SITE(pv_mmu_ops, read_cr2);
	SITE(pv_mmu_ops, read_cr3);
	SITE(pv_mmu_ops, write_cr3);
	SITE(pv_cpu_ops, clts);
	SITE(pv_cpu_ops, read_tsc);
100 101 102
#undef SITE

	patch_site:
103
		ret = paravirt_patch_insns(ibuf, len, start, end);
104 105 106
		break;

	default:
107
		ret = paravirt_patch_default(type, clobbers, ibuf, addr, len);
108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
		break;
	}

	return ret;
}

unsigned paravirt_patch_nop(void)
{
	return 0;
}

unsigned paravirt_patch_ignore(unsigned len)
{
	return len;
}

124 125 126 127 128
struct branch {
	unsigned char opcode;
	u32 delta;
} __attribute__((packed));

129 130 131
unsigned paravirt_patch_call(void *insnbuf,
			     const void *target, u16 tgt_clobbers,
			     unsigned long addr, u16 site_clobbers,
132 133
			     unsigned len)
{
134 135
	struct branch *b = insnbuf;
	unsigned long delta = (unsigned long)target - (addr+5);
136 137 138 139 140

	if (tgt_clobbers & ~site_clobbers)
		return len;	/* target would clobber too much for this site */
	if (len < 5)
		return len;	/* call too long for patch site */
141

142 143 144
	b->opcode = 0xe8; /* call */
	b->delta = delta;
	BUILD_BUG_ON(sizeof(*b) != 5);
145

146 147 148
	return 5;
}

149
unsigned paravirt_patch_jmp(void *insnbuf, const void *target,
150
			    unsigned long addr, unsigned len)
151
{
152 153
	struct branch *b = insnbuf;
	unsigned long delta = (unsigned long)target - (addr+5);
154 155 156 157

	if (len < 5)
		return len;	/* call too long for patch site */

158 159
	b->opcode = 0xe9;	/* jmp */
	b->delta = delta;
160 161 162 163

	return 5;
}

164 165 166 167 168 169 170 171 172 173 174 175 176 177 178
/* Neat trick to map patch type back to the call within the
 * corresponding structure. */
static void *get_call_destination(u8 type)
{
	struct paravirt_patch_template tmpl = {
		.pv_init_ops = pv_init_ops,
		.pv_time_ops = pv_time_ops,
		.pv_cpu_ops = pv_cpu_ops,
		.pv_irq_ops = pv_irq_ops,
		.pv_apic_ops = pv_apic_ops,
		.pv_mmu_ops = pv_mmu_ops,
	};
	return *((void **)&tmpl + type);
}

179 180
unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf,
				unsigned long addr, unsigned len)
181
{
182
	void *opfunc = get_call_destination(type);
183 184 185 186
	unsigned ret;

	if (opfunc == NULL)
		/* If there's no function, patch it with a ud2a (BUG) */
187
		ret = paravirt_patch_insns(insnbuf, len, ud2a, ud2a+sizeof(ud2a));
188 189 190
	else if (opfunc == paravirt_nop)
		/* If the operation is a nop, then nop the callsite */
		ret = paravirt_patch_nop();
191
	else if (type == PARAVIRT_PATCH(pv_cpu_ops.iret) ||
192
		 type == PARAVIRT_PATCH(pv_cpu_ops.irq_enable_syscall_ret))
193
		/* If operation requires a jmp, then jmp */
194
		ret = paravirt_patch_jmp(insnbuf, opfunc, addr, len);
195 196 197
	else
		/* Otherwise call the function; assume target could
		   clobber any caller-save reg */
198 199
		ret = paravirt_patch_call(insnbuf, opfunc, CLBR_ANY,
					  addr, clobbers, len);
200 201 202 203

	return ret;
}

204
unsigned paravirt_patch_insns(void *insnbuf, unsigned len,
205 206 207
			      const char *start, const char *end)
{
	unsigned insn_len = end - start;
208

209 210 211
	if (insn_len > len || start == NULL)
		insn_len = len;
	else
212
		memcpy(insnbuf, start, insn_len);
213 214 215 216

	return insn_len;
}

217 218
void init_IRQ(void)
{
219
	pv_irq_ops.init_IRQ();
220 221
}

222
static void native_flush_tlb(void)
223 224 225 226 227 228 229 230
{
	__native_flush_tlb();
}

/*
 * Global pages have to be flushed a bit differently. Not a real
 * performance problem because this does not happen often.
 */
231
static void native_flush_tlb_global(void)
232 233 234 235
{
	__native_flush_tlb_global();
}

236
static void native_flush_tlb_single(unsigned long addr)
237 238 239 240
{
	__native_flush_tlb_single(addr);
}

241
/* These are in entry.S */
242
extern void native_iret(void);
243
extern void native_irq_enable_syscall_ret(void);
244 245 246

static int __init print_banner(void)
{
247
	pv_init_ops.banner();
248 249 250 251
	return 0;
}
core_initcall(print_banner);

252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286
static struct resource reserve_ioports = {
	.start = 0,
	.end = IO_SPACE_LIMIT,
	.name = "paravirt-ioport",
	.flags = IORESOURCE_IO | IORESOURCE_BUSY,
};

static struct resource reserve_iomem = {
	.start = 0,
	.end = -1,
	.name = "paravirt-iomem",
	.flags = IORESOURCE_MEM | IORESOURCE_BUSY,
};

/*
 * Reserve the whole legacy IO space to prevent any legacy drivers
 * from wasting time probing for their hardware.  This is a fairly
 * brute-force approach to disabling all non-virtual drivers.
 *
 * Note that this must be called very early to have any effect.
 */
int paravirt_disable_iospace(void)
{
	int ret;

	ret = request_resource(&ioport_resource, &reserve_ioports);
	if (ret == 0) {
		ret = request_resource(&iomem_resource, &reserve_iomem);
		if (ret)
			release_resource(&reserve_ioports);
	}

	return ret;
}

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
static DEFINE_PER_CPU(enum paravirt_lazy_mode, paravirt_lazy_mode) = PARAVIRT_LAZY_NONE;

static inline void enter_lazy(enum paravirt_lazy_mode mode)
{
	BUG_ON(x86_read_percpu(paravirt_lazy_mode) != PARAVIRT_LAZY_NONE);
	BUG_ON(preemptible());

	x86_write_percpu(paravirt_lazy_mode, mode);
}

void paravirt_leave_lazy(enum paravirt_lazy_mode mode)
{
	BUG_ON(x86_read_percpu(paravirt_lazy_mode) != mode);
	BUG_ON(preemptible());

	x86_write_percpu(paravirt_lazy_mode, PARAVIRT_LAZY_NONE);
}

void paravirt_enter_lazy_mmu(void)
{
	enter_lazy(PARAVIRT_LAZY_MMU);
}

void paravirt_leave_lazy_mmu(void)
{
	paravirt_leave_lazy(PARAVIRT_LAZY_MMU);
}

void paravirt_enter_lazy_cpu(void)
{
	enter_lazy(PARAVIRT_LAZY_CPU);
}

void paravirt_leave_lazy_cpu(void)
{
	paravirt_leave_lazy(PARAVIRT_LAZY_CPU);
}

enum paravirt_lazy_mode paravirt_get_lazy_mode(void)
{
	return x86_read_percpu(paravirt_lazy_mode);
}

330
struct pv_info pv_info = {
331 332 333
	.name = "bare hardware",
	.paravirt_enabled = 0,
	.kernel_rpl = 0,
334
	.shared_kernel_pmd = 1,	/* Only used when CONFIG_X86_PAE is set */
335
};
336

337 338
struct pv_init_ops pv_init_ops = {
	.patch = native_patch,
339
	.banner = default_banner,
340
	.arch_setup = paravirt_nop,
341
	.memory_setup = machine_specific_memory_setup,
342 343 344 345
};

struct pv_time_ops pv_time_ops = {
	.time_init = hpet_time_init,
346 347
	.get_wallclock = native_get_wallclock,
	.set_wallclock = native_set_wallclock,
348 349 350 351 352
	.sched_clock = native_sched_clock,
	.get_cpu_khz = native_calculate_cpu_khz,
};

struct pv_irq_ops pv_irq_ops = {
353
	.init_IRQ = native_init_IRQ,
354 355 356 357 358 359 360
	.save_fl = native_save_fl,
	.restore_fl = native_restore_fl,
	.irq_disable = native_irq_disable,
	.irq_enable = native_irq_enable,
	.safe_halt = native_safe_halt,
	.halt = native_halt,
};
361

362
struct pv_cpu_ops pv_cpu_ops = {
363 364 365 366 367 368 369 370 371 372
	.cpuid = native_cpuid,
	.get_debugreg = native_get_debugreg,
	.set_debugreg = native_set_debugreg,
	.clts = native_clts,
	.read_cr0 = native_read_cr0,
	.write_cr0 = native_write_cr0,
	.read_cr4 = native_read_cr4,
	.read_cr4_safe = native_read_cr4_safe,
	.write_cr4 = native_write_cr4,
	.wbinvd = native_wbinvd,
373 374
	.read_msr = native_read_msr_safe,
	.write_msr = native_write_msr_safe,
375 376
	.read_tsc = native_read_tsc,
	.read_pmc = native_read_pmc,
377
	.read_tscp = native_read_tscp,
378 379 380 381 382 383 384 385
	.load_tr_desc = native_load_tr_desc,
	.set_ldt = native_set_ldt,
	.load_gdt = native_load_gdt,
	.load_idt = native_load_idt,
	.store_gdt = native_store_gdt,
	.store_idt = native_store_idt,
	.store_tr = native_store_tr,
	.load_tls = native_load_tls,
386
	.write_ldt_entry = native_write_ldt_entry,
387
	.write_gdt_entry = native_write_gdt_entry,
388
	.write_idt_entry = native_write_idt_entry,
389
	.load_sp0 = native_load_sp0,
390

391
	.irq_enable_syscall_ret = native_irq_enable_syscall_ret,
392 393
	.iret = native_iret,

394 395
	.set_iopl_mask = native_set_iopl_mask,
	.io_delay = native_io_delay,
396 397 398 399 400

	.lazy_mode = {
		.enter = paravirt_nop,
		.leave = paravirt_nop,
	},
401
};
402

403
struct pv_apic_ops pv_apic_ops = {
404 405 406 407
#ifdef CONFIG_X86_LOCAL_APIC
	.apic_write = native_apic_write,
	.apic_write_atomic = native_apic_write_atomic,
	.apic_read = native_apic_read,
Z
Zachary Amsden 已提交
408 409
	.setup_boot_clock = setup_boot_APIC_clock,
	.setup_secondary_clock = setup_secondary_APIC_clock,
410
	.startup_ipi_hook = paravirt_nop,
411
#endif
412 413 414
};

struct pv_mmu_ops pv_mmu_ops = {
415 416 417
	.pagetable_setup_start = native_pagetable_setup_start,
	.pagetable_setup_done = native_pagetable_setup_done,

418 419 420 421 422
	.read_cr2 = native_read_cr2,
	.write_cr2 = native_write_cr2,
	.read_cr3 = native_read_cr3,
	.write_cr3 = native_write_cr3,

423 424 425
	.flush_tlb_user = native_flush_tlb,
	.flush_tlb_kernel = native_flush_tlb_global,
	.flush_tlb_single = native_flush_tlb_single,
426
	.flush_tlb_others = native_flush_tlb_others,
427

428 429 430 431 432
	.alloc_pt = paravirt_nop,
	.alloc_pd = paravirt_nop,
	.alloc_pd_clone = paravirt_nop,
	.release_pt = paravirt_nop,
	.release_pd = paravirt_nop,
433

434 435 436
	.set_pte = native_set_pte,
	.set_pte_at = native_set_pte_at,
	.set_pmd = native_set_pmd,
437 438
	.pte_update = paravirt_nop,
	.pte_update_defer = paravirt_nop,
439

440 441 442 443
#ifdef CONFIG_HIGHPTE
	.kmap_atomic_pte = kmap_atomic,
#endif

444 445 446 447 448 449
#ifdef CONFIG_X86_PAE
	.set_pte_atomic = native_set_pte_atomic,
	.set_pte_present = native_set_pte_present,
	.set_pud = native_set_pud,
	.pte_clear = native_pte_clear,
	.pmd_clear = native_pmd_clear,
450 451 452

	.pmd_val = native_pmd_val,
	.make_pmd = native_make_pmd,
453 454
#endif

455 456 457 458 459 460
	.pte_val = native_pte_val,
	.pgd_val = native_pgd_val,

	.make_pte = native_make_pte,
	.make_pgd = native_make_pgd,

461 462 463
	.dup_mmap = paravirt_nop,
	.exit_mmap = paravirt_nop,
	.activate_mm = paravirt_nop,
464 465 466 467 468

	.lazy_mode = {
		.enter = paravirt_nop,
		.leave = paravirt_nop,
	},
469
};
470

471
EXPORT_SYMBOL_GPL(pv_time_ops);
472 473
EXPORT_SYMBOL    (pv_cpu_ops);
EXPORT_SYMBOL    (pv_mmu_ops);
474 475 476
EXPORT_SYMBOL_GPL(pv_apic_ops);
EXPORT_SYMBOL_GPL(pv_info);
EXPORT_SYMBOL    (pv_irq_ops);