paravirt.c 10.3 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
#define DEF_NATIVE(ops, name, code)					\
	extern const char start_##ops##_##name[], end_##ops##_##name[];	\
	asm("start_" #ops "_" #name ": " code "; end_" #ops "_" #name ":")

/* Undefined instruction for dealing with missing ops pointers. */
static const unsigned char ud2a[] = { 0x0f, 0x0b };
63

64 65 66 67 68 69 70 71 72 73
unsigned paravirt_patch_nop(void)
{
	return 0;
}

unsigned paravirt_patch_ignore(unsigned len)
{
	return len;
}

74 75 76 77 78
struct branch {
	unsigned char opcode;
	u32 delta;
} __attribute__((packed));

79 80 81
unsigned paravirt_patch_call(void *insnbuf,
			     const void *target, u16 tgt_clobbers,
			     unsigned long addr, u16 site_clobbers,
82 83
			     unsigned len)
{
84 85
	struct branch *b = insnbuf;
	unsigned long delta = (unsigned long)target - (addr+5);
86 87 88 89 90

	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 */
91

92 93 94
	b->opcode = 0xe8; /* call */
	b->delta = delta;
	BUILD_BUG_ON(sizeof(*b) != 5);
95

96 97 98
	return 5;
}

99
unsigned paravirt_patch_jmp(void *insnbuf, const void *target,
100
			    unsigned long addr, unsigned len)
101
{
102 103
	struct branch *b = insnbuf;
	unsigned long delta = (unsigned long)target - (addr+5);
104 105 106 107

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

108 109
	b->opcode = 0xe9;	/* jmp */
	b->delta = delta;
110 111 112 113

	return 5;
}

114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
/* 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);
}

129 130
unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf,
				unsigned long addr, unsigned len)
131
{
132
	void *opfunc = get_call_destination(type);
133 134 135 136
	unsigned ret;

	if (opfunc == NULL)
		/* If there's no function, patch it with a ud2a (BUG) */
137
		ret = paravirt_patch_insns(insnbuf, len, ud2a, ud2a+sizeof(ud2a));
138 139 140
	else if (opfunc == paravirt_nop)
		/* If the operation is a nop, then nop the callsite */
		ret = paravirt_patch_nop();
141
	else if (type == PARAVIRT_PATCH(pv_cpu_ops.iret) ||
142
		 type == PARAVIRT_PATCH(pv_cpu_ops.irq_enable_syscall_ret))
143
		/* If operation requires a jmp, then jmp */
144
		ret = paravirt_patch_jmp(insnbuf, opfunc, addr, len);
145 146 147
	else
		/* Otherwise call the function; assume target could
		   clobber any caller-save reg */
148 149
		ret = paravirt_patch_call(insnbuf, opfunc, CLBR_ANY,
					  addr, clobbers, len);
150 151 152 153

	return ret;
}

154
unsigned paravirt_patch_insns(void *insnbuf, unsigned len,
155 156 157
			      const char *start, const char *end)
{
	unsigned insn_len = end - start;
158

159 160 161
	if (insn_len > len || start == NULL)
		insn_len = len;
	else
162
		memcpy(insnbuf, start, insn_len);
163 164 165 166

	return insn_len;
}

167 168
void init_IRQ(void)
{
169
	pv_irq_ops.init_IRQ();
170 171
}

172
static void native_flush_tlb(void)
173 174 175 176 177 178 179 180
{
	__native_flush_tlb();
}

/*
 * Global pages have to be flushed a bit differently. Not a real
 * performance problem because this does not happen often.
 */
181
static void native_flush_tlb_global(void)
182 183 184 185
{
	__native_flush_tlb_global();
}

186
static void native_flush_tlb_single(unsigned long addr)
187 188 189 190
{
	__native_flush_tlb_single(addr);
}

191
/* These are in entry.S */
192
extern void native_iret(void);
193
extern void native_irq_enable_syscall_ret(void);
194 195 196

static int __init print_banner(void)
{
197
	pv_init_ops.banner();
198 199 200 201
	return 0;
}
core_initcall(print_banner);

202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
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;
}

237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
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);
}

280
struct pv_info pv_info = {
281 282 283
	.name = "bare hardware",
	.paravirt_enabled = 0,
	.kernel_rpl = 0,
284
	.shared_kernel_pmd = 1,	/* Only used when CONFIG_X86_PAE is set */
285
};
286

287 288
struct pv_init_ops pv_init_ops = {
	.patch = native_patch,
289
	.banner = default_banner,
290
	.arch_setup = paravirt_nop,
291
	.memory_setup = machine_specific_memory_setup,
292 293 294 295
};

struct pv_time_ops pv_time_ops = {
	.time_init = hpet_time_init,
296 297
	.get_wallclock = native_get_wallclock,
	.set_wallclock = native_set_wallclock,
298 299 300 301 302
	.sched_clock = native_sched_clock,
	.get_cpu_khz = native_calculate_cpu_khz,
};

struct pv_irq_ops pv_irq_ops = {
303
	.init_IRQ = native_init_IRQ,
304 305 306 307 308 309 310
	.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,
};
311

312
struct pv_cpu_ops pv_cpu_ops = {
313 314 315 316 317 318 319 320 321 322
	.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,
323 324
	.read_msr = native_read_msr_safe,
	.write_msr = native_write_msr_safe,
325 326
	.read_tsc = native_read_tsc,
	.read_pmc = native_read_pmc,
327
	.read_tscp = native_read_tscp,
328 329 330 331 332 333 334 335
	.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,
336
	.write_ldt_entry = native_write_ldt_entry,
337
	.write_gdt_entry = native_write_gdt_entry,
338
	.write_idt_entry = native_write_idt_entry,
339
	.load_sp0 = native_load_sp0,
340

341
	.irq_enable_syscall_ret = native_irq_enable_syscall_ret,
342
	.iret = native_iret,
343
	.swapgs = native_swapgs,
344

345 346
	.set_iopl_mask = native_set_iopl_mask,
	.io_delay = native_io_delay,
347 348 349 350 351

	.lazy_mode = {
		.enter = paravirt_nop,
		.leave = paravirt_nop,
	},
352
};
353

354
struct pv_apic_ops pv_apic_ops = {
355 356 357 358
#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
359 360
	.setup_boot_clock = setup_boot_APIC_clock,
	.setup_secondary_clock = setup_secondary_APIC_clock,
361
	.startup_ipi_hook = paravirt_nop,
362
#endif
363 364 365
};

struct pv_mmu_ops pv_mmu_ops = {
366 367 368
	.pagetable_setup_start = native_pagetable_setup_start,
	.pagetable_setup_done = native_pagetable_setup_done,

369 370 371 372 373
	.read_cr2 = native_read_cr2,
	.write_cr2 = native_write_cr2,
	.read_cr3 = native_read_cr3,
	.write_cr3 = native_write_cr3,

374 375 376
	.flush_tlb_user = native_flush_tlb,
	.flush_tlb_kernel = native_flush_tlb_global,
	.flush_tlb_single = native_flush_tlb_single,
377
	.flush_tlb_others = native_flush_tlb_others,
378

379 380 381 382 383
	.alloc_pt = paravirt_nop,
	.alloc_pd = paravirt_nop,
	.alloc_pd_clone = paravirt_nop,
	.release_pt = paravirt_nop,
	.release_pd = paravirt_nop,
384

385 386 387
	.set_pte = native_set_pte,
	.set_pte_at = native_set_pte_at,
	.set_pmd = native_set_pmd,
388 389
	.pte_update = paravirt_nop,
	.pte_update_defer = paravirt_nop,
390

391 392 393 394
#ifdef CONFIG_HIGHPTE
	.kmap_atomic_pte = kmap_atomic,
#endif

395 396 397 398 399 400
#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,
401 402 403

	.pmd_val = native_pmd_val,
	.make_pmd = native_make_pmd,
404 405
#endif

406 407 408 409 410 411
	.pte_val = native_pte_val,
	.pgd_val = native_pgd_val,

	.make_pte = native_make_pte,
	.make_pgd = native_make_pgd,

412 413 414
	.dup_mmap = paravirt_nop,
	.exit_mmap = paravirt_nop,
	.activate_mm = paravirt_nop,
415 416 417 418 419

	.lazy_mode = {
		.enter = paravirt_nop,
		.leave = paravirt_nop,
	},
420
};
421

422
EXPORT_SYMBOL_GPL(pv_time_ops);
423 424
EXPORT_SYMBOL    (pv_cpu_ops);
EXPORT_SYMBOL    (pv_mmu_ops);
425 426 427
EXPORT_SYMBOL_GPL(pv_apic_ops);
EXPORT_SYMBOL_GPL(pv_info);
EXPORT_SYMBOL    (pv_irq_ops);