[net-next-2.6.git] / arch / x86 / kernel / paravirt.c

/*  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

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

#include <linux/errno.h>
#include <linux/module.h>
#include <linux/efi.h>
#include <linux/bcd.h>
#include <linux/highmem.h>

#include <asm/bug.h>
#include <asm/paravirt.h>
#include <asm/desc.h>
#include <asm/setup.h>
#include <asm/pgtable.h>
#include <asm/time.h>
#include <asm/pgalloc.h>
#include <asm/irq.h>
#include <asm/delay.h>
#include <asm/fixmap.h>
#include <asm/apic.h>
#include <asm/tlbflush.h>
#include <asm/timer.h>

/* nop stub */
void _paravirt_nop(void)
{
}

/* identity function, which can be inlined */
u32 _paravirt_ident_32(u32 x)
{
	return x;
}

u64 _paravirt_ident_64(u64 x)
{
	return x;
}

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

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

/* Simple instruction patching code. */
#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 };

unsigned paravirt_patch_nop(void)
{
	return 0;
}

unsigned paravirt_patch_ignore(unsigned len)
{
	return len;
}

struct branch {
	unsigned char opcode;
	u32 delta;
} __attribute__((packed));

unsigned paravirt_patch_call(void *insnbuf,
			     const void *target, u16 tgt_clobbers,
			     unsigned long addr, u16 site_clobbers,
			     unsigned len)
{
	struct branch *b = insnbuf;
	unsigned long delta = (unsigned long)target - (addr+5);

	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 */

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

	return 5;
}

unsigned paravirt_patch_jmp(void *insnbuf, const void *target,
			    unsigned long addr, unsigned len)
{
	struct branch *b = insnbuf;
	unsigned long delta = (unsigned long)target - (addr+5);

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

	b->opcode = 0xe9;	/* jmp */
	b->delta = delta;

	return 5;
}

/* 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,
		.pv_lock_ops = pv_lock_ops,
	};
	return *((void **)&tmpl + type);
}

unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf,
				unsigned long addr, unsigned len)
{
	void *opfunc = get_call_destination(type);
	unsigned ret;

	if (opfunc == NULL)
		/* If there's no function, patch it with a ud2a (BUG) */
		ret = paravirt_patch_insns(insnbuf, len, ud2a, ud2a+sizeof(ud2a));
	else if (opfunc == _paravirt_nop)
		/* If the operation is a nop, then nop the callsite */
		ret = paravirt_patch_nop();

	/* identity functions just return their single argument */
	else if (opfunc == _paravirt_ident_32)
		ret = paravirt_patch_ident_32(insnbuf, len);
	else if (opfunc == _paravirt_ident_64)
		ret = paravirt_patch_ident_64(insnbuf, len);

	else if (type == PARAVIRT_PATCH(pv_cpu_ops.iret) ||
		 type == PARAVIRT_PATCH(pv_cpu_ops.irq_enable_sysexit) ||
		 type == PARAVIRT_PATCH(pv_cpu_ops.usergs_sysret32) ||
		 type == PARAVIRT_PATCH(pv_cpu_ops.usergs_sysret64))
		/* If operation requires a jmp, then jmp */
		ret = paravirt_patch_jmp(insnbuf, opfunc, addr, len);
	else
		/* Otherwise call the function; assume target could
		   clobber any caller-save reg */
		ret = paravirt_patch_call(insnbuf, opfunc, CLBR_ANY,
					  addr, clobbers, len);

	return ret;
}

unsigned paravirt_patch_insns(void *insnbuf, unsigned len,
			      const char *start, const char *end)
{
	unsigned insn_len = end - start;

	if (insn_len > len || start == NULL)
		insn_len = len;
	else
		memcpy(insnbuf, start, insn_len);

	return insn_len;
}

void init_IRQ(void)
{
	pv_irq_ops.init_IRQ();
}

static void native_flush_tlb(void)
{
	__native_flush_tlb();
}

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

static void native_flush_tlb_single(unsigned long addr)
{
	__native_flush_tlb_single(addr);
}

/* These are in entry.S */
extern void native_iret(void);
extern void native_irq_enable_sysexit(void);
extern void native_usergs_sysret32(void);
extern void native_usergs_sysret64(void);

static int __init print_banner(void)
{
	pv_init_ops.banner();
	return 0;
}
core_initcall(print_banner);

static struct resource reserve_ioports = {
	.start = 0,
	.end = IO_SPACE_LIMIT,
	.name = "paravirt-ioport",
	.flags = IORESOURCE_IO | 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)
{
	return request_resource(&ioport_resource, &reserve_ioports);
}

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(__get_cpu_var(paravirt_lazy_mode) != PARAVIRT_LAZY_NONE);
	BUG_ON(preemptible());

	__get_cpu_var(paravirt_lazy_mode) = mode;
}

static void leave_lazy(enum paravirt_lazy_mode mode)
{
	BUG_ON(__get_cpu_var(paravirt_lazy_mode) != mode);
	BUG_ON(preemptible());

	__get_cpu_var(paravirt_lazy_mode) = PARAVIRT_LAZY_NONE;
}

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

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

void paravirt_start_context_switch(struct task_struct *prev)
{
	if (percpu_read(paravirt_lazy_mode) == PARAVIRT_LAZY_MMU) {
		arch_leave_lazy_mmu_mode();
		set_ti_thread_flag(task_thread_info(prev), TIF_LAZY_MMU_UPDATES);
	}
	enter_lazy(PARAVIRT_LAZY_CPU);
}

void paravirt_end_context_switch(struct task_struct *next)
{
	leave_lazy(PARAVIRT_LAZY_CPU);

	if (test_and_clear_ti_thread_flag(task_thread_info(next), TIF_LAZY_MMU_UPDATES))
		arch_enter_lazy_mmu_mode();
}

enum paravirt_lazy_mode paravirt_get_lazy_mode(void)
{
	if (in_interrupt())
		return PARAVIRT_LAZY_NONE;

	return __get_cpu_var(paravirt_lazy_mode);
}

void arch_flush_lazy_mmu_mode(void)
{
	preempt_disable();

	if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_MMU) {
		WARN_ON(preempt_count() == 1);
		arch_leave_lazy_mmu_mode();
		arch_enter_lazy_mmu_mode();
	}

	preempt_enable();
}

struct pv_info pv_info = {
	.name = "bare hardware",
	.paravirt_enabled = 0,
	.kernel_rpl = 0,
	.shared_kernel_pmd = 1,	/* Only used when CONFIG_X86_PAE is set */
};

struct pv_init_ops pv_init_ops = {
	.patch = native_patch,
	.banner = default_banner,
	.arch_setup = paravirt_nop,
	.memory_setup = machine_specific_memory_setup,
};

struct pv_time_ops pv_time_ops = {
	.time_init = hpet_time_init,
	.get_wallclock = native_get_wallclock,
	.set_wallclock = native_set_wallclock,
	.sched_clock = native_sched_clock,
	.get_tsc_khz = native_calibrate_tsc,
};

struct pv_irq_ops pv_irq_ops = {
	.init_IRQ = native_init_IRQ,
	.save_fl = __PV_IS_CALLEE_SAVE(native_save_fl),
	.restore_fl = __PV_IS_CALLEE_SAVE(native_restore_fl),
	.irq_disable = __PV_IS_CALLEE_SAVE(native_irq_disable),
	.irq_enable = __PV_IS_CALLEE_SAVE(native_irq_enable),
	.safe_halt = native_safe_halt,
	.halt = native_halt,
#ifdef CONFIG_X86_64
	.adjust_exception_frame = paravirt_nop,
#endif
};

struct pv_cpu_ops pv_cpu_ops = {
	.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,
#ifdef CONFIG_X86_64
	.read_cr8 = native_read_cr8,
	.write_cr8 = native_write_cr8,
#endif
	.wbinvd = native_wbinvd,
	.read_msr = native_read_msr_safe,
	.read_msr_amd = native_read_msr_amd_safe,
	.write_msr = native_write_msr_safe,
	.read_tsc = native_read_tsc,
	.read_pmc = native_read_pmc,
	.read_tscp = native_read_tscp,
	.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,
#ifdef CONFIG_X86_64
	.load_gs_index = native_load_gs_index,
#endif
	.write_ldt_entry = native_write_ldt_entry,
	.write_gdt_entry = native_write_gdt_entry,
	.write_idt_entry = native_write_idt_entry,

	.alloc_ldt = paravirt_nop,
	.free_ldt = paravirt_nop,

	.load_sp0 = native_load_sp0,

#if defined(CONFIG_X86_32) || defined(CONFIG_IA32_EMULATION)
	.irq_enable_sysexit = native_irq_enable_sysexit,
#endif
#ifdef CONFIG_X86_64
#ifdef CONFIG_IA32_EMULATION
	.usergs_sysret32 = native_usergs_sysret32,
#endif
	.usergs_sysret64 = native_usergs_sysret64,
#endif
	.iret = native_iret,
	.swapgs = native_swapgs,

	.set_iopl_mask = native_set_iopl_mask,
	.io_delay = native_io_delay,

	.start_context_switch = paravirt_nop,
	.end_context_switch = paravirt_nop,
};

struct pv_apic_ops pv_apic_ops = {
#ifdef CONFIG_X86_LOCAL_APIC
	.setup_boot_clock = setup_boot_APIC_clock,
	.setup_secondary_clock = setup_secondary_APIC_clock,
	.startup_ipi_hook = paravirt_nop,
#endif
};

#if defined(CONFIG_X86_32) && !defined(CONFIG_X86_PAE)
/* 32-bit pagetable entries */
#define PTE_IDENT	__PV_IS_CALLEE_SAVE(_paravirt_ident_32)
#else
/* 64-bit pagetable entries */
#define PTE_IDENT	__PV_IS_CALLEE_SAVE(_paravirt_ident_64)
#endif

struct pv_mmu_ops pv_mmu_ops = {
#ifndef CONFIG_X86_64
	.pagetable_setup_start = native_pagetable_setup_start,
	.pagetable_setup_done = native_pagetable_setup_done,
#else
	.pagetable_setup_start = paravirt_nop,
	.pagetable_setup_done = paravirt_nop,
#endif

	.read_cr2 = native_read_cr2,
	.write_cr2 = native_write_cr2,
	.read_cr3 = native_read_cr3,
	.write_cr3 = native_write_cr3,

	.flush_tlb_user = native_flush_tlb,
	.flush_tlb_kernel = native_flush_tlb_global,
	.flush_tlb_single = native_flush_tlb_single,
	.flush_tlb_others = native_flush_tlb_others,

	.pgd_alloc = __paravirt_pgd_alloc,
	.pgd_free = paravirt_nop,

	.alloc_pte = paravirt_nop,
	.alloc_pmd = paravirt_nop,
	.alloc_pmd_clone = paravirt_nop,
	.alloc_pud = paravirt_nop,
	.release_pte = paravirt_nop,
	.release_pmd = paravirt_nop,
	.release_pud = paravirt_nop,

	.set_pte = native_set_pte,
	.set_pte_at = native_set_pte_at,
	.set_pmd = native_set_pmd,
	.pte_update = paravirt_nop,
	.pte_update_defer = paravirt_nop,

	.ptep_modify_prot_start = __ptep_modify_prot_start,
	.ptep_modify_prot_commit = __ptep_modify_prot_commit,

#ifdef CONFIG_HIGHPTE
	.kmap_atomic_pte = kmap_atomic,
#endif

#if PAGETABLE_LEVELS >= 3
#ifdef CONFIG_X86_PAE
	.set_pte_atomic = native_set_pte_atomic,
	.set_pte_present = native_set_pte_present,
	.pte_clear = native_pte_clear,
	.pmd_clear = native_pmd_clear,
#endif
	.set_pud = native_set_pud,

	.pmd_val = PTE_IDENT,
	.make_pmd = PTE_IDENT,

#if PAGETABLE_LEVELS == 4
	.pud_val = PTE_IDENT,
	.make_pud = PTE_IDENT,

	.set_pgd = native_set_pgd,
#endif
#endif /* PAGETABLE_LEVELS >= 3 */

	.pte_val = PTE_IDENT,
	.pgd_val = PTE_IDENT,

	.make_pte = PTE_IDENT,
	.make_pgd = PTE_IDENT,

	.dup_mmap = paravirt_nop,
	.exit_mmap = paravirt_nop,
	.activate_mm = paravirt_nop,

	.lazy_mode = {
		.enter = paravirt_nop,
		.leave = paravirt_nop,
	},

	.set_fixmap = native_set_fixmap,
};

EXPORT_SYMBOL_GPL(pv_time_ops);
EXPORT_SYMBOL    (pv_cpu_ops);
EXPORT_SYMBOL    (pv_mmu_ops);
EXPORT_SYMBOL_GPL(pv_apic_ops);
EXPORT_SYMBOL_GPL(pv_info);
EXPORT_SYMBOL    (pv_irq_ops);
Commit	Line	Data
d3561b7f RR	1	/* Paravirtualization interfaces
	2	Copyright (C) 2006 Rusty Russell IBM Corporation
	3
	4	This program is free software; you can redistribute it and/or modify
	5	it under the terms of the GNU General Public License as published by
	6	the Free Software Foundation; either version 2 of the License, or
	7	(at your option) any later version.
	8
	9	This program is distributed in the hope that it will be useful,
	10	but WITHOUT ANY WARRANTY; without even the implied warranty of
	11	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
	12	GNU General Public License for more details.
	13
	14	You should have received a copy of the GNU General Public License
	15	along with this program; if not, write to the Free Software
	16	Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
b1df07bd GOC	17
b1df07bd GOC	18	2007 - x86_64 support added by Glauber de Oliveira Costa, Red Hat Inc
d3561b7f	19	*/
b1df07bd	20
d3561b7f RR	21	#include <linux/errno.h>
	22	#include <linux/module.h>
	23	#include <linux/efi.h>
	24	#include <linux/bcd.h>
ce6234b5	25	#include <linux/highmem.h>
d3561b7f RR	26
	27	#include <asm/bug.h>
	28	#include <asm/paravirt.h>
	29	#include <asm/desc.h>
	30	#include <asm/setup.h>
a312b37b	31	#include <asm/pgtable.h>
d3561b7f	32	#include <asm/time.h>
eba0045f	33	#include <asm/pgalloc.h>
d3561b7f RR	34	#include <asm/irq.h>
d3561b7f RR	35	#include <asm/delay.h>
13623d79 RR	36	#include <asm/fixmap.h>
13623d79 RR	37	#include <asm/apic.h>
da181a8b	38	#include <asm/tlbflush.h>
6cb9a835	39	#include <asm/timer.h>
d3561b7f RR	40
d3561b7f RR	41	/* nop stub */
45876233	42	void _paravirt_nop(void)
d3561b7f RR	43	{
	44	}
	45
41edafdb JF	46	/* identity function, which can be inlined */
	47	u32 _paravirt_ident_32(u32 x)
	48	{
	49	return x;
	50	}
	51
	52	u64 _paravirt_ident_64(u64 x)
	53	{
	54	return x;
	55	}
	56
d3561b7f RR	57	static void __init default_banner(void)
	58	{
	59	printk(KERN_INFO "Booting paravirtualized kernel on %s\n",
93b1eab3	60	pv_info.name);
d3561b7f RR	61	}
	62
	63	char *memory_setup(void)
	64	{
93b1eab3	65	return pv_init_ops.memory_setup();
d3561b7f RR	66	}
d3561b7f RR	67
139ec7c4	68	/* Simple instruction patching code. */
93b1eab3 JF	69	#define DEF_NATIVE(ops, name, code) \
	70	extern const char start_##ops##_##name[], end_##ops##_##name[]; \
	71	asm("start_" #ops "_" #name ": " code "; end_" #ops "_" #name ":")
	72
93b1eab3 JF	73	/* Undefined instruction for dealing with missing ops pointers. */
93b1eab3 JF	74	static const unsigned char ud2a[] = { 0x0f, 0x0b };
139ec7c4	75
63f70270 JF	76	unsigned paravirt_patch_nop(void)
	77	{
	78	return 0;
	79	}
	80
	81	unsigned paravirt_patch_ignore(unsigned len)
	82	{
	83	return len;
	84	}
	85
19d36ccd AK	86	struct branch {
	87	unsigned char opcode;
	88	u32 delta;
	89	} __attribute__((packed));
	90
ab144f5e AK	91	unsigned paravirt_patch_call(void *insnbuf,
	92	const void *target, u16 tgt_clobbers,
	93	unsigned long addr, u16 site_clobbers,
63f70270 JF	94	unsigned len)
63f70270 JF	95	{
ab144f5e AK	96	struct branch *b = insnbuf;
ab144f5e AK	97	unsigned long delta = (unsigned long)target - (addr+5);
63f70270 JF	98
	99	if (tgt_clobbers & ~site_clobbers)
	100	return len; /* target would clobber too much for this site */
	101	if (len < 5)
	102	return len; /* call too long for patch site */
139ec7c4	103
ab144f5e AK	104	b->opcode = 0xe8; /* call */
	105	b->delta = delta;
	106	BUILD_BUG_ON(sizeof(*b) != 5);
139ec7c4	107
63f70270 JF	108	return 5;
	109	}
	110
93b1eab3	111	unsigned paravirt_patch_jmp(void insnbuf, const void target,
ab144f5e	112	unsigned long addr, unsigned len)
63f70270	113	{
ab144f5e AK	114	struct branch *b = insnbuf;
ab144f5e AK	115	unsigned long delta = (unsigned long)target - (addr+5);
63f70270 JF	116
	117	if (len < 5)
	118	return len; /* call too long for patch site */
	119
ab144f5e AK	120	b->opcode = 0xe9; /* jmp */
ab144f5e AK	121	b->delta = delta;
63f70270 JF	122
	123	return 5;
	124	}
	125
93b1eab3 JF	126	/* Neat trick to map patch type back to the call within the
	127	* corresponding structure. */
	128	static void *get_call_destination(u8 type)
	129	{
	130	struct paravirt_patch_template tmpl = {
	131	.pv_init_ops = pv_init_ops,
93b1eab3 JF	132	.pv_time_ops = pv_time_ops,
	133	.pv_cpu_ops = pv_cpu_ops,
	134	.pv_irq_ops = pv_irq_ops,
	135	.pv_apic_ops = pv_apic_ops,
	136	.pv_mmu_ops = pv_mmu_ops,
74d4affd	137	.pv_lock_ops = pv_lock_ops,
93b1eab3 JF	138	};
	139	return ((void *)&tmpl + type);
	140	}
	141
ab144f5e AK	142	unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf,
ab144f5e AK	143	unsigned long addr, unsigned len)
63f70270	144	{
93b1eab3	145	void *opfunc = get_call_destination(type);
63f70270 JF	146	unsigned ret;
	147
	148	if (opfunc == NULL)
	149	/* If there's no function, patch it with a ud2a (BUG) */
93b1eab3	150	ret = paravirt_patch_insns(insnbuf, len, ud2a, ud2a+sizeof(ud2a));
41edafdb	151	else if (opfunc == _paravirt_nop)
63f70270 JF	152	/* If the operation is a nop, then nop the callsite */
63f70270 JF	153	ret = paravirt_patch_nop();
41edafdb JF	154
	155	/* identity functions just return their single argument */
	156	else if (opfunc == _paravirt_ident_32)
	157	ret = paravirt_patch_ident_32(insnbuf, len);
	158	else if (opfunc == _paravirt_ident_64)
	159	ret = paravirt_patch_ident_64(insnbuf, len);
	160
93b1eab3	161	else if (type == PARAVIRT_PATCH(pv_cpu_ops.iret) \|\|
d75cd22f	162	type == PARAVIRT_PATCH(pv_cpu_ops.irq_enable_sysexit) \|\|
2be29982 JF	163	type == PARAVIRT_PATCH(pv_cpu_ops.usergs_sysret32) \|\|
2be29982 JF	164	type == PARAVIRT_PATCH(pv_cpu_ops.usergs_sysret64))
63f70270	165	/* If operation requires a jmp, then jmp */
93b1eab3	166	ret = paravirt_patch_jmp(insnbuf, opfunc, addr, len);
63f70270 JF	167	else
	168	/* Otherwise call the function; assume target could
	169	clobber any caller-save reg */
ab144f5e AK	170	ret = paravirt_patch_call(insnbuf, opfunc, CLBR_ANY,
ab144f5e AK	171	addr, clobbers, len);
63f70270 JF	172
	173	return ret;
	174	}
	175
ab144f5e	176	unsigned paravirt_patch_insns(void *insnbuf, unsigned len,
63f70270 JF	177	const char start, const char end)
	178	{
	179	unsigned insn_len = end - start;
139ec7c4	180
63f70270 JF	181	if (insn_len > len \|\| start == NULL)
	182	insn_len = len;
	183	else
ab144f5e	184	memcpy(insnbuf, start, insn_len);
139ec7c4	185
139ec7c4 RR	186	return insn_len;
	187	}
	188
d3561b7f RR	189	void init_IRQ(void)
d3561b7f RR	190	{
93b1eab3	191	pv_irq_ops.init_IRQ();
d3561b7f RR	192	}
d3561b7f RR	193
1a1eecd1	194	static void native_flush_tlb(void)
da181a8b RR	195	{
	196	__native_flush_tlb();
	197	}
	198
	199	/*
	200	* Global pages have to be flushed a bit differently. Not a real
	201	* performance problem because this does not happen often.
	202	*/
1a1eecd1	203	static void native_flush_tlb_global(void)
da181a8b RR	204	{
	205	__native_flush_tlb_global();
	206	}
	207
63f70270	208	static void native_flush_tlb_single(unsigned long addr)
da181a8b RR	209	{
	210	__native_flush_tlb_single(addr);
	211	}
	212
d3561b7f	213	/* These are in entry.S */
1a1eecd1	214	extern void native_iret(void);
d75cd22f	215	extern void native_irq_enable_sysexit(void);
2be29982 JF	216	extern void native_usergs_sysret32(void);
2be29982 JF	217	extern void native_usergs_sysret64(void);
d3561b7f RR	218
	219	static int __init print_banner(void)
	220	{
93b1eab3	221	pv_init_ops.banner();
d3561b7f RR	222	return 0;
	223	}
	224	core_initcall(print_banner);
	225
d572929c JF	226	static struct resource reserve_ioports = {
	227	.start = 0,
	228	.end = IO_SPACE_LIMIT,
	229	.name = "paravirt-ioport",
	230	.flags = IORESOURCE_IO \| IORESOURCE_BUSY,
	231	};
	232
d572929c JF	233	/*
	234	* Reserve the whole legacy IO space to prevent any legacy drivers
	235	* from wasting time probing for their hardware. This is a fairly
	236	* brute-force approach to disabling all non-virtual drivers.
	237	*
	238	* Note that this must be called very early to have any effect.
	239	*/
	240	int paravirt_disable_iospace(void)
	241	{
f7743fe6	242	return request_resource(&ioport_resource, &reserve_ioports);
d572929c JF	243	}
d572929c JF	244
8965c1c0 JF	245	static DEFINE_PER_CPU(enum paravirt_lazy_mode, paravirt_lazy_mode) = PARAVIRT_LAZY_NONE;
	246
	247	static inline void enter_lazy(enum paravirt_lazy_mode mode)
	248	{
bfd074e0	249	BUG_ON(__get_cpu_var(paravirt_lazy_mode) != PARAVIRT_LAZY_NONE);
8965c1c0 JF	250	BUG_ON(preemptible());
8965c1c0 JF	251
bfd074e0	252	__get_cpu_var(paravirt_lazy_mode) = mode;
8965c1c0 JF	253	}
8965c1c0 JF	254
b407fc57	255	static void leave_lazy(enum paravirt_lazy_mode mode)
8965c1c0	256	{
bfd074e0	257	BUG_ON(__get_cpu_var(paravirt_lazy_mode) != mode);
8965c1c0 JF	258	BUG_ON(preemptible());
8965c1c0 JF	259
bfd074e0	260	__get_cpu_var(paravirt_lazy_mode) = PARAVIRT_LAZY_NONE;
8965c1c0 JF	261	}
	262
	263	void paravirt_enter_lazy_mmu(void)
	264	{
	265	enter_lazy(PARAVIRT_LAZY_MMU);
	266	}
	267
	268	void paravirt_leave_lazy_mmu(void)
	269	{
b407fc57	270	leave_lazy(PARAVIRT_LAZY_MMU);
8965c1c0 JF	271	}
8965c1c0 JF	272
224101ed	273	void paravirt_start_context_switch(struct task_struct *prev)
8965c1c0	274	{
b407fc57 JF	275	if (percpu_read(paravirt_lazy_mode) == PARAVIRT_LAZY_MMU) {
b407fc57 JF	276	arch_leave_lazy_mmu_mode();
224101ed	277	set_ti_thread_flag(task_thread_info(prev), TIF_LAZY_MMU_UPDATES);
b407fc57	278	}
8965c1c0 JF	279	enter_lazy(PARAVIRT_LAZY_CPU);
	280	}
	281
224101ed	282	void paravirt_end_context_switch(struct task_struct *next)
8965c1c0	283	{
b407fc57 JF	284	leave_lazy(PARAVIRT_LAZY_CPU);
b407fc57 JF	285
224101ed	286	if (test_and_clear_ti_thread_flag(task_thread_info(next), TIF_LAZY_MMU_UPDATES))
b407fc57	287	arch_enter_lazy_mmu_mode();
8965c1c0 JF	288	}
	289
	290	enum paravirt_lazy_mode paravirt_get_lazy_mode(void)
	291	{
b8bcfe99 JF	292	if (in_interrupt())
	293	return PARAVIRT_LAZY_NONE;
	294
bfd074e0	295	return __get_cpu_var(paravirt_lazy_mode);
8965c1c0 JF	296	}
8965c1c0 JF	297
d85cf93d JF	298	void arch_flush_lazy_mmu_mode(void)
	299	{
	300	preempt_disable();
	301
	302	if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_MMU) {
34b0900d	303	WARN_ON(preempt_count() == 1);
d85cf93d JF	304	arch_leave_lazy_mmu_mode();
	305	arch_enter_lazy_mmu_mode();
	306	}
	307
	308	preempt_enable();
	309	}
	310
93b1eab3	311	struct pv_info pv_info = {
d3561b7f RR	312	.name = "bare hardware",
	313	.paravirt_enabled = 0,
	314	.kernel_rpl = 0,
5311ab62	315	.shared_kernel_pmd = 1, /* Only used when CONFIG_X86_PAE is set */
93b1eab3	316	};
d3561b7f	317
93b1eab3 JF	318	struct pv_init_ops pv_init_ops = {
93b1eab3 JF	319	.patch = native_patch,
d3561b7f	320	.banner = default_banner,
45876233	321	.arch_setup = paravirt_nop,
d3561b7f	322	.memory_setup = machine_specific_memory_setup,
93b1eab3 JF	323	};
	324
	325	struct pv_time_ops pv_time_ops = {
	326	.time_init = hpet_time_init,
d3561b7f RR	327	.get_wallclock = native_get_wallclock,
d3561b7f RR	328	.set_wallclock = native_set_wallclock,
93b1eab3	329	.sched_clock = native_sched_clock,
e93ef949	330	.get_tsc_khz = native_calibrate_tsc,
93b1eab3 JF	331	};
	332
	333	struct pv_irq_ops pv_irq_ops = {
d3561b7f	334	.init_IRQ = native_init_IRQ,
ecb93d1c JF	335	.save_fl = __PV_IS_CALLEE_SAVE(native_save_fl),
	336	.restore_fl = __PV_IS_CALLEE_SAVE(native_restore_fl),
	337	.irq_disable = __PV_IS_CALLEE_SAVE(native_irq_disable),
	338	.irq_enable = __PV_IS_CALLEE_SAVE(native_irq_enable),
93b1eab3 JF	339	.safe_halt = native_safe_halt,
93b1eab3 JF	340	.halt = native_halt,
fab58420 JF	341	#ifdef CONFIG_X86_64
	342	.adjust_exception_frame = paravirt_nop,
	343	#endif
93b1eab3	344	};
d3561b7f	345
93b1eab3	346	struct pv_cpu_ops pv_cpu_ops = {
d3561b7f RR	347	.cpuid = native_cpuid,
	348	.get_debugreg = native_get_debugreg,
	349	.set_debugreg = native_set_debugreg,
	350	.clts = native_clts,
	351	.read_cr0 = native_read_cr0,
	352	.write_cr0 = native_write_cr0,
d3561b7f RR	353	.read_cr4 = native_read_cr4,
	354	.read_cr4_safe = native_read_cr4_safe,
	355	.write_cr4 = native_write_cr4,
88b4755f GOC	356	#ifdef CONFIG_X86_64
	357	.read_cr8 = native_read_cr8,
	358	.write_cr8 = native_write_cr8,
	359	#endif
d3561b7f	360	.wbinvd = native_wbinvd,
90a0a06a	361	.read_msr = native_read_msr_safe,
b05f78f5	362	.read_msr_amd = native_read_msr_amd_safe,
90a0a06a	363	.write_msr = native_write_msr_safe,
d3561b7f RR	364	.read_tsc = native_read_tsc,
d3561b7f RR	365	.read_pmc = native_read_pmc,
e5aaac44	366	.read_tscp = native_read_tscp,
d3561b7f RR	367	.load_tr_desc = native_load_tr_desc,
	368	.set_ldt = native_set_ldt,
	369	.load_gdt = native_load_gdt,
	370	.load_idt = native_load_idt,
	371	.store_gdt = native_store_gdt,
	372	.store_idt = native_store_idt,
	373	.store_tr = native_store_tr,
	374	.load_tls = native_load_tls,
9f9d489a JF	375	#ifdef CONFIG_X86_64
	376	.load_gs_index = native_load_gs_index,
	377	#endif
75b8bb3e	378	.write_ldt_entry = native_write_ldt_entry,
014b15be	379	.write_gdt_entry = native_write_gdt_entry,
8d947344	380	.write_idt_entry = native_write_idt_entry,
38ffbe66 JF	381
	382	.alloc_ldt = paravirt_nop,
	383	.free_ldt = paravirt_nop,
	384
faca6227	385	.load_sp0 = native_load_sp0,
d3561b7f	386
102d0a4b	387	#if defined(CONFIG_X86_32) \|\| defined(CONFIG_IA32_EMULATION)
d75cd22f	388	.irq_enable_sysexit = native_irq_enable_sysexit,
102d0a4b	389	#endif
2be29982	390	#ifdef CONFIG_X86_64
102d0a4b	391	#ifdef CONFIG_IA32_EMULATION
2be29982	392	.usergs_sysret32 = native_usergs_sysret32,
102d0a4b	393	#endif
2be29982	394	.usergs_sysret64 = native_usergs_sysret64,
d75cd22f	395	#endif
93b1eab3	396	.iret = native_iret,
e801f864	397	.swapgs = native_swapgs,
93b1eab3	398
d3561b7f RR	399	.set_iopl_mask = native_set_iopl_mask,
d3561b7f RR	400	.io_delay = native_io_delay,
8965c1c0	401
224101ed JF	402	.start_context_switch = paravirt_nop,
224101ed JF	403	.end_context_switch = paravirt_nop,
93b1eab3	404	};
d3561b7f	405
93b1eab3	406	struct pv_apic_ops pv_apic_ops = {
13623d79	407	#ifdef CONFIG_X86_LOCAL_APIC
bbab4f3b ZA	408	.setup_boot_clock = setup_boot_APIC_clock,
bbab4f3b ZA	409	.setup_secondary_clock = setup_secondary_APIC_clock,
0260c196	410	.startup_ipi_hook = paravirt_nop,
13623d79	411	#endif
93b1eab3 JF	412	};
93b1eab3 JF	413
41edafdb JF	414	#if defined(CONFIG_X86_32) && !defined(CONFIG_X86_PAE)
41edafdb JF	415	/* 32-bit pagetable entries */
da5de7c2	416	#define PTE_IDENT __PV_IS_CALLEE_SAVE(_paravirt_ident_32)
41edafdb JF	417	#else
41edafdb JF	418	/* 64-bit pagetable entries */
da5de7c2	419	#define PTE_IDENT __PV_IS_CALLEE_SAVE(_paravirt_ident_64)
41edafdb JF	420	#endif
41edafdb JF	421
93b1eab3	422	struct pv_mmu_ops pv_mmu_ops = {
d8dd8eec	423	#ifndef CONFIG_X86_64
b239fb25 JF	424	.pagetable_setup_start = native_pagetable_setup_start,
b239fb25 JF	425	.pagetable_setup_done = native_pagetable_setup_done,
a312b37b EH	426	#else
	427	.pagetable_setup_start = paravirt_nop,
	428	.pagetable_setup_done = paravirt_nop,
d8dd8eec	429	#endif
b239fb25	430
93b1eab3 JF	431	.read_cr2 = native_read_cr2,
	432	.write_cr2 = native_write_cr2,
	433	.read_cr3 = native_read_cr3,
	434	.write_cr3 = native_write_cr3,
	435
da181a8b RR	436	.flush_tlb_user = native_flush_tlb,
	437	.flush_tlb_kernel = native_flush_tlb_global,
	438	.flush_tlb_single = native_flush_tlb_single,
d4c10477	439	.flush_tlb_others = native_flush_tlb_others,
da181a8b	440
eba0045f JF	441	.pgd_alloc = __paravirt_pgd_alloc,
	442	.pgd_free = paravirt_nop,
	443
6944a9c8 JF	444	.alloc_pte = paravirt_nop,
	445	.alloc_pmd = paravirt_nop,
	446	.alloc_pmd_clone = paravirt_nop,
2761fa09	447	.alloc_pud = paravirt_nop,
6944a9c8 JF	448	.release_pte = paravirt_nop,
6944a9c8 JF	449	.release_pmd = paravirt_nop,
2761fa09	450	.release_pud = paravirt_nop,
c119ecce	451
da181a8b RR	452	.set_pte = native_set_pte,
	453	.set_pte_at = native_set_pte_at,
	454	.set_pmd = native_set_pmd,
45876233 JF	455	.pte_update = paravirt_nop,
45876233 JF	456	.pte_update_defer = paravirt_nop,
3dc494e8	457
08b882c6 JF	458	.ptep_modify_prot_start = __ptep_modify_prot_start,
	459	.ptep_modify_prot_commit = __ptep_modify_prot_commit,
	460
ce6234b5 JF	461	#ifdef CONFIG_HIGHPTE
	462	.kmap_atomic_pte = kmap_atomic,
	463	#endif
	464
f95f2f7b	465	#if PAGETABLE_LEVELS >= 3
da181a8b RR	466	#ifdef CONFIG_X86_PAE
	467	.set_pte_atomic = native_set_pte_atomic,
	468	.set_pte_present = native_set_pte_present,
da181a8b RR	469	.pte_clear = native_pte_clear,
da181a8b RR	470	.pmd_clear = native_pmd_clear,
f95f2f7b EH	471	#endif
f95f2f7b EH	472	.set_pud = native_set_pud,
da5de7c2 JF	473
	474	.pmd_val = PTE_IDENT,
	475	.make_pmd = PTE_IDENT,
f95f2f7b EH	476
f95f2f7b EH	477	#if PAGETABLE_LEVELS == 4
da5de7c2 JF	478	.pud_val = PTE_IDENT,
	479	.make_pud = PTE_IDENT,
	480
f95f2f7b	481	.set_pgd = native_set_pgd,
da181a8b	482	#endif
f95f2f7b	483	#endif /* PAGETABLE_LEVELS >= 3 */
da181a8b	484
da5de7c2 JF	485	.pte_val = PTE_IDENT,
da5de7c2 JF	486	.pgd_val = PTE_IDENT,
3dc494e8	487
da5de7c2 JF	488	.make_pte = PTE_IDENT,
da5de7c2 JF	489	.make_pgd = PTE_IDENT,
3dc494e8	490
d6dd61c8 JF	491	.dup_mmap = paravirt_nop,
	492	.exit_mmap = paravirt_nop,
	493	.activate_mm = paravirt_nop,
8965c1c0 JF	494
	495	.lazy_mode = {
	496	.enter = paravirt_nop,
	497	.leave = paravirt_nop,
	498	},
aeaaa59c JF	499
aeaaa59c JF	500	.set_fixmap = native_set_fixmap,
d3561b7f	501	};
0dbe5a11	502
93b1eab3	503	EXPORT_SYMBOL_GPL(pv_time_ops);
f97b8954 JF	504	EXPORT_SYMBOL (pv_cpu_ops);
f97b8954 JF	505	EXPORT_SYMBOL (pv_mmu_ops);
93b1eab3 JF	506	EXPORT_SYMBOL_GPL(pv_apic_ops);
	507	EXPORT_SYMBOL_GPL(pv_info);
	508	EXPORT_SYMBOL (pv_irq_ops);