machine_kexec_32.c 6.2 KB
Newer Older
1
/*
Dave Jones's avatar
Dave Jones committed
2
 * handle transition of Linux booting another kernel
3 4 5 6 7 8 9 10 11
 * Copyright (C) 2002-2005 Eric Biederman  <ebiederm@xmission.com>
 *
 * This source code is licensed under the GNU General Public License,
 * Version 2.  See the file COPYING for more details.
 */

#include <linux/mm.h>
#include <linux/kexec.h>
#include <linux/delay.h>
12
#include <linux/init.h>
Ken'ichi Ohmichi's avatar
Ken'ichi Ohmichi committed
13
#include <linux/numa.h>
Ingo Molnar's avatar
Ingo Molnar committed
14
#include <linux/ftrace.h>
Huang Ying's avatar
Huang Ying committed
15
#include <linux/suspend.h>
16
#include <linux/gfp.h>
Ingo Molnar's avatar
Ingo Molnar committed
17

18 19 20 21 22 23 24
#include <asm/pgtable.h>
#include <asm/pgalloc.h>
#include <asm/tlbflush.h>
#include <asm/mmu_context.h>
#include <asm/io.h>
#include <asm/apic.h>
#include <asm/cpufeature.h>
25
#include <asm/desc.h>
26
#include <asm/system.h>
Huang Ying's avatar
Huang Ying committed
27
#include <asm/cacheflush.h>
28 29 30

static void set_idt(void *newidt, __u16 limit)
{
31
	struct desc_ptr curidt;
32 33

	/* ia32 supports unaliged loads & stores */
34 35
	curidt.size    = limit;
	curidt.address = (unsigned long)newidt;
36

37
	load_idt(&curidt);
38
}
39 40 41 42


static void set_gdt(void *newgdt, __u16 limit)
{
43
	struct desc_ptr curgdt;
44 45

	/* ia32 supports unaligned loads & stores */
46 47
	curgdt.size    = limit;
	curgdt.address = (unsigned long)newgdt;
48

49
	load_gdt(&curgdt);
50
}
51 52 53 54 55 56 57 58 59

static void load_segments(void)
{
#define __STR(X) #X
#define STR(X) __STR(X)

	__asm__ __volatile__ (
		"\tljmp $"STR(__KERNEL_CS)",$1f\n"
		"\t1:\n"
Michael Matz's avatar
Michael Matz committed
60 61 62 63 64 65 66
		"\tmovl $"STR(__KERNEL_DS)",%%eax\n"
		"\tmovl %%eax,%%ds\n"
		"\tmovl %%eax,%%es\n"
		"\tmovl %%eax,%%fs\n"
		"\tmovl %%eax,%%gs\n"
		"\tmovl %%eax,%%ss\n"
		::: "eax", "memory");
67 68 69 70
#undef STR
#undef __STR
}

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
static void machine_kexec_free_page_tables(struct kimage *image)
{
	free_page((unsigned long)image->arch.pgd);
#ifdef CONFIG_X86_PAE
	free_page((unsigned long)image->arch.pmd0);
	free_page((unsigned long)image->arch.pmd1);
#endif
	free_page((unsigned long)image->arch.pte0);
	free_page((unsigned long)image->arch.pte1);
}

static int machine_kexec_alloc_page_tables(struct kimage *image)
{
	image->arch.pgd = (pgd_t *)get_zeroed_page(GFP_KERNEL);
#ifdef CONFIG_X86_PAE
	image->arch.pmd0 = (pmd_t *)get_zeroed_page(GFP_KERNEL);
	image->arch.pmd1 = (pmd_t *)get_zeroed_page(GFP_KERNEL);
#endif
	image->arch.pte0 = (pte_t *)get_zeroed_page(GFP_KERNEL);
	image->arch.pte1 = (pte_t *)get_zeroed_page(GFP_KERNEL);
	if (!image->arch.pgd ||
#ifdef CONFIG_X86_PAE
	    !image->arch.pmd0 || !image->arch.pmd1 ||
#endif
	    !image->arch.pte0 || !image->arch.pte1) {
		machine_kexec_free_page_tables(image);
		return -ENOMEM;
	}
	return 0;
}

102 103 104
/*
 * A architecture hook called to validate the
 * proposed image and prepare the control pages
105
 * as needed.  The pages for KEXEC_CONTROL_PAGE_SIZE
106 107 108 109 110 111 112
 * have been allocated, but the segments have yet
 * been copied into the kernel.
 *
 * Do what every setup is needed on image and the
 * reboot code buffer to allow us to avoid allocations
 * later.
 *
113 114
 * - Make control page executable.
 * - Allocate page tables
115 116 117
 */
int machine_kexec_prepare(struct kimage *image)
{
Huang Ying's avatar
Huang Ying committed
118 119
	if (nx_enabled)
		set_pages_x(image->control_code_page, 1);
120
	return machine_kexec_alloc_page_tables(image);
121 122 123 124 125 126 127 128
}

/*
 * Undo anything leftover by machine_kexec_prepare
 * when an image is freed.
 */
void machine_kexec_cleanup(struct kimage *image)
{
Huang Ying's avatar
Huang Ying committed
129 130
	if (nx_enabled)
		set_pages_nx(image->control_code_page, 1);
131
	machine_kexec_free_page_tables(image);
132 133 134 135 136 137
}

/*
 * Do not allocate memory (or fail in any way) in machine_kexec().
 * We are past the point of no return, committed to rebooting now.
 */
Huang Ying's avatar
Huang Ying committed
138
void machine_kexec(struct kimage *image)
139
{
140 141
	unsigned long page_list[PAGES_NR];
	void *control_page;
Huang Ying's avatar
Huang Ying committed
142
	int save_ftrace_enabled;
Huang Ying's avatar
Huang Ying committed
143 144 145 146 147 148
	asmlinkage unsigned long
		(*relocate_kernel_ptr)(unsigned long indirection_page,
				       unsigned long control_page,
				       unsigned long start_address,
				       unsigned int has_pae,
				       unsigned int preserve_context);
149

Huang Ying's avatar
Huang Ying committed
150 151 152 153 154 155
#ifdef CONFIG_KEXEC_JUMP
	if (kexec_image->preserve_context)
		save_processor_state();
#endif

	save_ftrace_enabled = __ftrace_enabled_save();
Ingo Molnar's avatar
Ingo Molnar committed
156

157 158 159
	/* Interrupts aren't acceptable while we reboot */
	local_irq_disable();

160 161 162 163 164 165 166 167 168 169 170 171
	if (image->preserve_context) {
#ifdef CONFIG_X86_IO_APIC
		/* We need to put APICs in legacy mode so that we can
		 * get timer interrupts in second kernel. kexec/kdump
		 * paths already have calls to disable_IO_APIC() in
		 * one form or other. kexec jump path also need
		 * one.
		 */
		disable_IO_APIC();
#endif
	}

172
	control_page = page_address(image->control_code_page);
173
	memcpy(control_page, relocate_kernel, KEXEC_CONTROL_CODE_MAX_SIZE);
174

Huang Ying's avatar
Huang Ying committed
175
	relocate_kernel_ptr = control_page;
176
	page_list[PA_CONTROL_PAGE] = __pa(control_page);
Huang Ying's avatar
Huang Ying committed
177
	page_list[VA_CONTROL_PAGE] = (unsigned long)control_page;
178 179
	page_list[PA_PGD] = __pa(image->arch.pgd);
	page_list[VA_PGD] = (unsigned long)image->arch.pgd;
180
#ifdef CONFIG_X86_PAE
181 182 183 184
	page_list[PA_PMD_0] = __pa(image->arch.pmd0);
	page_list[VA_PMD_0] = (unsigned long)image->arch.pmd0;
	page_list[PA_PMD_1] = __pa(image->arch.pmd1);
	page_list[VA_PMD_1] = (unsigned long)image->arch.pmd1;
185
#endif
186 187 188 189
	page_list[PA_PTE_0] = __pa(image->arch.pte0);
	page_list[VA_PTE_0] = (unsigned long)image->arch.pte0;
	page_list[PA_PTE_1] = __pa(image->arch.pte1);
	page_list[VA_PTE_1] = (unsigned long)image->arch.pte1;
190 191 192 193

	if (image->type == KEXEC_TYPE_DEFAULT)
		page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page)
						<< PAGE_SHIFT);
194

195 196 197 198 199
	/* The segment registers are funny things, they have both a
	 * visible and an invisible part.  Whenever the visible part is
	 * set to a specific selector, the invisible part is loaded
	 * with from a table in memory.  At no other time is the
	 * descriptor table in memory accessed.
200 201 202 203 204 205 206 207 208 209 210 211
	 *
	 * I take advantage of this here by force loading the
	 * segments, before I zap the gdt with an invalid value.
	 */
	load_segments();
	/* The gdt & idt are now invalid.
	 * If you want to load them you must set up your own idt & gdt.
	 */
	set_gdt(phys_to_virt(0),0);
	set_idt(phys_to_virt(0),0);

	/* now call it */
Huang Ying's avatar
Huang Ying committed
212 213 214 215
	image->start = relocate_kernel_ptr((unsigned long)image->head,
					   (unsigned long)page_list,
					   image->start, cpu_has_pae,
					   image->preserve_context);
Huang Ying's avatar
Huang Ying committed
216 217 218 219 220 221 222

#ifdef CONFIG_KEXEC_JUMP
	if (kexec_image->preserve_context)
		restore_processor_state();
#endif

	__ftrace_enabled_restore(save_ftrace_enabled);
223
}
224

Ken'ichi Ohmichi's avatar
Ken'ichi Ohmichi committed
225 226
void arch_crash_save_vmcoreinfo(void)
{
227
#ifdef CONFIG_NUMA
228 229
	VMCOREINFO_SYMBOL(node_data);
	VMCOREINFO_LENGTH(node_data, MAX_NUMNODES);
Ken'ichi Ohmichi's avatar
Ken'ichi Ohmichi committed
230 231
#endif
#ifdef CONFIG_X86_PAE
232
	VMCOREINFO_CONFIG(X86_PAE);
Ken'ichi Ohmichi's avatar
Ken'ichi Ohmichi committed
233 234 235
#endif
}