[linux-2.6-block.git] / arch / x86 / kernel / machine_kexec_64.c

/*
 * handle transition of Linux booting another kernel
 * 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.
 */

#define pr_fmt(fmt)	"kexec: " fmt

#include <linux/mm.h>
#include <linux/kexec.h>
#include <linux/string.h>
#include <linux/gfp.h>
#include <linux/reboot.h>
#include <linux/numa.h>
#include <linux/ftrace.h>
#include <linux/io.h>
#include <linux/suspend.h>
#include <linux/vmalloc.h>

#include <asm/init.h>
#include <asm/pgtable.h>
#include <asm/tlbflush.h>
#include <asm/mmu_context.h>
#include <asm/io_apic.h>
#include <asm/debugreg.h>
#include <asm/kexec-bzimage64.h>
#include <asm/setup.h>
#include <asm/set_memory.h>

#ifdef CONFIG_KEXEC_FILE
static struct kexec_file_ops *kexec_file_loaders[] = {
		&kexec_bzImage64_ops,
};
#endif

static void free_transition_pgtable(struct kimage *image)
{
	free_page((unsigned long)image->arch.p4d);
	free_page((unsigned long)image->arch.pud);
	free_page((unsigned long)image->arch.pmd);
	free_page((unsigned long)image->arch.pte);
}

static int init_transition_pgtable(struct kimage *image, pgd_t *pgd)
{
	p4d_t *p4d;
	pud_t *pud;
	pmd_t *pmd;
	pte_t *pte;
	unsigned long vaddr, paddr;
	int result = -ENOMEM;

	vaddr = (unsigned long)relocate_kernel;
	paddr = __pa(page_address(image->control_code_page)+PAGE_SIZE);
	pgd += pgd_index(vaddr);
	if (!pgd_present(*pgd)) {
		p4d = (p4d_t *)get_zeroed_page(GFP_KERNEL);
		if (!p4d)
			goto err;
		image->arch.p4d = p4d;
		set_pgd(pgd, __pgd(__pa(p4d) | _KERNPG_TABLE));
	}
	p4d = p4d_offset(pgd, vaddr);
	if (!p4d_present(*p4d)) {
		pud = (pud_t *)get_zeroed_page(GFP_KERNEL);
		if (!pud)
			goto err;
		image->arch.pud = pud;
		set_p4d(p4d, __p4d(__pa(pud) | _KERNPG_TABLE));
	}
	pud = pud_offset(p4d, vaddr);
	if (!pud_present(*pud)) {
		pmd = (pmd_t *)get_zeroed_page(GFP_KERNEL);
		if (!pmd)
			goto err;
		image->arch.pmd = pmd;
		set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE));
	}
	pmd = pmd_offset(pud, vaddr);
	if (!pmd_present(*pmd)) {
		pte = (pte_t *)get_zeroed_page(GFP_KERNEL);
		if (!pte)
			goto err;
		image->arch.pte = pte;
		set_pmd(pmd, __pmd(__pa(pte) | _KERNPG_TABLE));
	}
	pte = pte_offset_kernel(pmd, vaddr);
	set_pte(pte, pfn_pte(paddr >> PAGE_SHIFT, PAGE_KERNEL_EXEC_NOENC));
	return 0;
err:
	free_transition_pgtable(image);
	return result;
}

static void *alloc_pgt_page(void *data)
{
	struct kimage *image = (struct kimage *)data;
	struct page *page;
	void *p = NULL;

	page = kimage_alloc_control_pages(image, 0);
	if (page) {
		p = page_address(page);
		clear_page(p);
	}

	return p;
}

static int init_pgtable(struct kimage *image, unsigned long start_pgtable)
{
	struct x86_mapping_info info = {
		.alloc_pgt_page	= alloc_pgt_page,
		.context	= image,
		.page_flag	= __PAGE_KERNEL_LARGE_EXEC,
		.kernpg_flag	= _KERNPG_TABLE_NOENC,
	};
	unsigned long mstart, mend;
	pgd_t *level4p;
	int result;
	int i;

	level4p = (pgd_t *)__va(start_pgtable);
	clear_page(level4p);

	if (direct_gbpages)
		info.direct_gbpages = true;

	for (i = 0; i < nr_pfn_mapped; i++) {
		mstart = pfn_mapped[i].start << PAGE_SHIFT;
		mend   = pfn_mapped[i].end << PAGE_SHIFT;

		result = kernel_ident_mapping_init(&info,
						 level4p, mstart, mend);
		if (result)
			return result;
	}

	/*
	 * segments's mem ranges could be outside 0 ~ max_pfn,
	 * for example when jump back to original kernel from kexeced kernel.
	 * or first kernel is booted with user mem map, and second kernel
	 * could be loaded out of that range.
	 */
	for (i = 0; i < image->nr_segments; i++) {
		mstart = image->segment[i].mem;
		mend   = mstart + image->segment[i].memsz;

		result = kernel_ident_mapping_init(&info,
						 level4p, mstart, mend);

		if (result)
			return result;
	}

	return init_transition_pgtable(image, level4p);
}

static void set_idt(void *newidt, u16 limit)
{
	struct desc_ptr curidt;

	/* x86-64 supports unaliged loads & stores */
	curidt.size    = limit;
	curidt.address = (unsigned long)newidt;

	__asm__ __volatile__ (
		"lidtq %0\n"
		: : "m" (curidt)
		);
};


static void set_gdt(void *newgdt, u16 limit)
{
	struct desc_ptr curgdt;

	/* x86-64 supports unaligned loads & stores */
	curgdt.size    = limit;
	curgdt.address = (unsigned long)newgdt;

	__asm__ __volatile__ (
		"lgdtq %0\n"
		: : "m" (curgdt)
		);
};

static void load_segments(void)
{
	__asm__ __volatile__ (
		"\tmovl %0,%%ds\n"
		"\tmovl %0,%%es\n"
		"\tmovl %0,%%ss\n"
		"\tmovl %0,%%fs\n"
		"\tmovl %0,%%gs\n"
		: : "a" (__KERNEL_DS) : "memory"
		);
}

#ifdef CONFIG_KEXEC_FILE
/* Update purgatory as needed after various image segments have been prepared */
static int arch_update_purgatory(struct kimage *image)
{
	int ret = 0;

	if (!image->file_mode)
		return 0;

	/* Setup copying of backup region */
	if (image->type == KEXEC_TYPE_CRASH) {
		ret = kexec_purgatory_get_set_symbol(image,
				"purgatory_backup_dest",
				&image->arch.backup_load_addr,
				sizeof(image->arch.backup_load_addr), 0);
		if (ret)
			return ret;

		ret = kexec_purgatory_get_set_symbol(image,
				"purgatory_backup_src",
				&image->arch.backup_src_start,
				sizeof(image->arch.backup_src_start), 0);
		if (ret)
			return ret;

		ret = kexec_purgatory_get_set_symbol(image,
				"purgatory_backup_sz",
				&image->arch.backup_src_sz,
				sizeof(image->arch.backup_src_sz), 0);
		if (ret)
			return ret;
	}

	return ret;
}
#else /* !CONFIG_KEXEC_FILE */
static inline int arch_update_purgatory(struct kimage *image)
{
	return 0;
}
#endif /* CONFIG_KEXEC_FILE */

int machine_kexec_prepare(struct kimage *image)
{
	unsigned long start_pgtable;
	int result;

	/* Calculate the offsets */
	start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT;

	/* Setup the identity mapped 64bit page table */
	result = init_pgtable(image, start_pgtable);
	if (result)
		return result;

	/* update purgatory as needed */
	result = arch_update_purgatory(image);
	if (result)
		return result;

	return 0;
}

void machine_kexec_cleanup(struct kimage *image)
{
	free_transition_pgtable(image);
}

/*
 * Do not allocate memory (or fail in any way) in machine_kexec().
 * We are past the point of no return, committed to rebooting now.
 */
void machine_kexec(struct kimage *image)
{
	unsigned long page_list[PAGES_NR];
	void *control_page;
	int save_ftrace_enabled;

#ifdef CONFIG_KEXEC_JUMP
	if (image->preserve_context)
		save_processor_state();
#endif

	save_ftrace_enabled = __ftrace_enabled_save();

	/* Interrupts aren't acceptable while we reboot */
	local_irq_disable();
	hw_breakpoint_disable();

	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
	}

	control_page = page_address(image->control_code_page) + PAGE_SIZE;
	memcpy(control_page, relocate_kernel, KEXEC_CONTROL_CODE_MAX_SIZE);

	page_list[PA_CONTROL_PAGE] = virt_to_phys(control_page);
	page_list[VA_CONTROL_PAGE] = (unsigned long)control_page;
	page_list[PA_TABLE_PAGE] =
	  (unsigned long)__pa(page_address(image->control_code_page));

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

	/*
	 * 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.
	 *
	 * 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 */
	image->start = relocate_kernel((unsigned long)image->head,
				       (unsigned long)page_list,
				       image->start,
				       image->preserve_context,
				       sme_active());

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

	__ftrace_enabled_restore(save_ftrace_enabled);
}

void arch_crash_save_vmcoreinfo(void)
{
	VMCOREINFO_NUMBER(phys_base);
	VMCOREINFO_SYMBOL(init_top_pgt);

#ifdef CONFIG_NUMA
	VMCOREINFO_SYMBOL(node_data);
	VMCOREINFO_LENGTH(node_data, MAX_NUMNODES);
#endif
	vmcoreinfo_append_str("KERNELOFFSET=%lx\n",
			      kaslr_offset());
	VMCOREINFO_NUMBER(KERNEL_IMAGE_SIZE);
}

/* arch-dependent functionality related to kexec file-based syscall */

#ifdef CONFIG_KEXEC_FILE
int arch_kexec_kernel_image_probe(struct kimage *image, void *buf,
				  unsigned long buf_len)
{
	int i, ret = -ENOEXEC;
	struct kexec_file_ops *fops;

	for (i = 0; i < ARRAY_SIZE(kexec_file_loaders); i++) {
		fops = kexec_file_loaders[i];
		if (!fops || !fops->probe)
			continue;

		ret = fops->probe(buf, buf_len);
		if (!ret) {
			image->fops = fops;
			return ret;
		}
	}

	return ret;
}

void *arch_kexec_kernel_image_load(struct kimage *image)
{
	vfree(image->arch.elf_headers);
	image->arch.elf_headers = NULL;

	if (!image->fops || !image->fops->load)
		return ERR_PTR(-ENOEXEC);

	return image->fops->load(image, image->kernel_buf,
				 image->kernel_buf_len, image->initrd_buf,
				 image->initrd_buf_len, image->cmdline_buf,
				 image->cmdline_buf_len);
}

int arch_kimage_file_post_load_cleanup(struct kimage *image)
{
	if (!image->fops || !image->fops->cleanup)
		return 0;

	return image->fops->cleanup(image->image_loader_data);
}

#ifdef CONFIG_KEXEC_VERIFY_SIG
int arch_kexec_kernel_verify_sig(struct kimage *image, void *kernel,
				 unsigned long kernel_len)
{
	if (!image->fops || !image->fops->verify_sig) {
		pr_debug("kernel loader does not support signature verification.");
		return -EKEYREJECTED;
	}

	return image->fops->verify_sig(kernel, kernel_len);
}
#endif

/*
 * Apply purgatory relocations.
 *
 * ehdr: Pointer to elf headers
 * sechdrs: Pointer to section headers.
 * relsec: section index of SHT_RELA section.
 *
 * TODO: Some of the code belongs to generic code. Move that in kexec.c.
 */
int arch_kexec_apply_relocations_add(const Elf64_Ehdr *ehdr,
				     Elf64_Shdr *sechdrs, unsigned int relsec)
{
	unsigned int i;
	Elf64_Rela *rel;
	Elf64_Sym *sym;
	void *location;
	Elf64_Shdr *section, *symtabsec;
	unsigned long address, sec_base, value;
	const char *strtab, *name, *shstrtab;

	/*
	 * ->sh_offset has been modified to keep the pointer to section
	 * contents in memory
	 */
	rel = (void *)sechdrs[relsec].sh_offset;

	/* Section to which relocations apply */
	section = &sechdrs[sechdrs[relsec].sh_info];

	pr_debug("Applying relocate section %u to %u\n", relsec,
		 sechdrs[relsec].sh_info);

	/* Associated symbol table */
	symtabsec = &sechdrs[sechdrs[relsec].sh_link];

	/* String table */
	if (symtabsec->sh_link >= ehdr->e_shnum) {
		/* Invalid strtab section number */
		pr_err("Invalid string table section index %d\n",
		       symtabsec->sh_link);
		return -ENOEXEC;
	}

	strtab = (char *)sechdrs[symtabsec->sh_link].sh_offset;

	/* section header string table */
	shstrtab = (char *)sechdrs[ehdr->e_shstrndx].sh_offset;

	for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rel); i++) {

		/*
		 * rel[i].r_offset contains byte offset from beginning
		 * of section to the storage unit affected.
		 *
		 * This is location to update (->sh_offset). This is temporary
		 * buffer where section is currently loaded. This will finally
		 * be loaded to a different address later, pointed to by
		 * ->sh_addr. kexec takes care of moving it
		 *  (kexec_load_segment()).
		 */
		location = (void *)(section->sh_offset + rel[i].r_offset);

		/* Final address of the location */
		address = section->sh_addr + rel[i].r_offset;

		/*
		 * rel[i].r_info contains information about symbol table index
		 * w.r.t which relocation must be made and type of relocation
		 * to apply. ELF64_R_SYM() and ELF64_R_TYPE() macros get
		 * these respectively.
		 */
		sym = (Elf64_Sym *)symtabsec->sh_offset +
				ELF64_R_SYM(rel[i].r_info);

		if (sym->st_name)
			name = strtab + sym->st_name;
		else
			name = shstrtab + sechdrs[sym->st_shndx].sh_name;

		pr_debug("Symbol: %s info: %02x shndx: %02x value=%llx size: %llx\n",
			 name, sym->st_info, sym->st_shndx, sym->st_value,
			 sym->st_size);

		if (sym->st_shndx == SHN_UNDEF) {
			pr_err("Undefined symbol: %s\n", name);
			return -ENOEXEC;
		}

		if (sym->st_shndx == SHN_COMMON) {
			pr_err("symbol '%s' in common section\n", name);
			return -ENOEXEC;
		}

		if (sym->st_shndx == SHN_ABS)
			sec_base = 0;
		else if (sym->st_shndx >= ehdr->e_shnum) {
			pr_err("Invalid section %d for symbol %s\n",
			       sym->st_shndx, name);
			return -ENOEXEC;
		} else
			sec_base = sechdrs[sym->st_shndx].sh_addr;

		value = sym->st_value;
		value += sec_base;
		value += rel[i].r_addend;

		switch (ELF64_R_TYPE(rel[i].r_info)) {
		case R_X86_64_NONE:
			break;
		case R_X86_64_64:
			*(u64 *)location = value;
			break;
		case R_X86_64_32:
			*(u32 *)location = value;
			if (value != *(u32 *)location)
				goto overflow;
			break;
		case R_X86_64_32S:
			*(s32 *)location = value;
			if ((s64)value != *(s32 *)location)
				goto overflow;
			break;
		case R_X86_64_PC32:
			value -= (u64)address;
			*(u32 *)location = value;
			break;
		default:
			pr_err("Unknown rela relocation: %llu\n",
			       ELF64_R_TYPE(rel[i].r_info));
			return -ENOEXEC;
		}
	}
	return 0;

overflow:
	pr_err("Overflow in relocation type %d value 0x%lx\n",
	       (int)ELF64_R_TYPE(rel[i].r_info), value);
	return -ENOEXEC;
}
#endif /* CONFIG_KEXEC_FILE */

static int
kexec_mark_range(unsigned long start, unsigned long end, bool protect)
{
	struct page *page;
	unsigned int nr_pages;

	/*
	 * For physical range: [start, end]. We must skip the unassigned
	 * crashk resource with zero-valued "end" member.
	 */
	if (!end || start > end)
		return 0;

	page = pfn_to_page(start >> PAGE_SHIFT);
	nr_pages = (end >> PAGE_SHIFT) - (start >> PAGE_SHIFT) + 1;
	if (protect)
		return set_pages_ro(page, nr_pages);
	else
		return set_pages_rw(page, nr_pages);
}

static void kexec_mark_crashkres(bool protect)
{
	unsigned long control;

	kexec_mark_range(crashk_low_res.start, crashk_low_res.end, protect);

	/* Don't touch the control code page used in crash_kexec().*/
	control = PFN_PHYS(page_to_pfn(kexec_crash_image->control_code_page));
	/* Control code page is located in the 2nd page. */
	kexec_mark_range(crashk_res.start, control + PAGE_SIZE - 1, protect);
	control += KEXEC_CONTROL_PAGE_SIZE;
	kexec_mark_range(control, crashk_res.end, protect);
}

void arch_kexec_protect_crashkres(void)
{
	kexec_mark_crashkres(true);
}

void arch_kexec_unprotect_crashkres(void)
{
	kexec_mark_crashkres(false);
}

int arch_kexec_post_alloc_pages(void *vaddr, unsigned int pages, gfp_t gfp)
{
	/*
	 * If SME is active we need to be sure that kexec pages are
	 * not encrypted because when we boot to the new kernel the
	 * pages won't be accessed encrypted (initially).
	 */
	return set_memory_decrypted((unsigned long)vaddr, pages);
}

void arch_kexec_pre_free_pages(void *vaddr, unsigned int pages)
{
	/*
	 * If SME is active we need to reset the pages back to being
	 * an encrypted mapping before freeing them.
	 */
	set_memory_encrypted((unsigned long)vaddr, pages);
}
Commit	Line	Data
5234f5eb	1	/*
835c34a1	2	* handle transition of Linux booting another kernel
5234f5eb EB	3	* Copyright (C) 2002-2005 Eric Biederman <ebiederm@xmission.com>
	4	*
	5	* This source code is licensed under the GNU General Public License,
	6	* Version 2. See the file COPYING for more details.
	7	*/
	8
12db5562 VG	9	#define pr_fmt(fmt) "kexec: " fmt
12db5562 VG	10
5234f5eb EB	11	#include <linux/mm.h>
5234f5eb EB	12	#include <linux/kexec.h>
5234f5eb	13	#include <linux/string.h>
5a0e3ad6	14	#include <linux/gfp.h>
5234f5eb	15	#include <linux/reboot.h>
fd59d231	16	#include <linux/numa.h>
f43fdad8	17	#include <linux/ftrace.h>
fef3a7a1	18	#include <linux/io.h>
fee7b0d8	19	#include <linux/suspend.h>
d6472302	20	#include <linux/vmalloc.h>
f43fdad8	21
9ebdc79f	22	#include <asm/init.h>
5234f5eb	23	#include <asm/pgtable.h>
5234f5eb EB	24	#include <asm/tlbflush.h>
5234f5eb EB	25	#include <asm/mmu_context.h>
8643e28d	26	#include <asm/io_apic.h>
17f557e5	27	#include <asm/debugreg.h>
27f48d3e	28	#include <asm/kexec-bzimage64.h>
4545c898	29	#include <asm/setup.h>
d1163651	30	#include <asm/set_memory.h>
8bf27556	31
74ca317c	32	#ifdef CONFIG_KEXEC_FILE
cb105258	33	static struct kexec_file_ops *kexec_file_loaders[] = {
27f48d3e	34	&kexec_bzImage64_ops,
cb105258	35	};
74ca317c	36	#endif
cb105258	37
f5deb796 HY	38	static void free_transition_pgtable(struct kimage *image)
f5deb796 HY	39	{
7f689041	40	free_page((unsigned long)image->arch.p4d);
f5deb796 HY	41	free_page((unsigned long)image->arch.pud);
	42	free_page((unsigned long)image->arch.pmd);
	43	free_page((unsigned long)image->arch.pte);
	44	}
	45
	46	static int init_transition_pgtable(struct kimage image, pgd_t pgd)
	47	{
7f689041	48	p4d_t *p4d;
f5deb796 HY	49	pud_t *pud;
	50	pmd_t *pmd;
	51	pte_t *pte;
	52	unsigned long vaddr, paddr;
	53	int result = -ENOMEM;
	54
	55	vaddr = (unsigned long)relocate_kernel;
	56	paddr = __pa(page_address(image->control_code_page)+PAGE_SIZE);
	57	pgd += pgd_index(vaddr);
	58	if (!pgd_present(*pgd)) {
7f689041 KS	59	p4d = (p4d_t *)get_zeroed_page(GFP_KERNEL);
	60	if (!p4d)
	61	goto err;
	62	image->arch.p4d = p4d;
	63	set_pgd(pgd, __pgd(__pa(p4d) \| _KERNPG_TABLE));
	64	}
	65	p4d = p4d_offset(pgd, vaddr);
	66	if (!p4d_present(*p4d)) {
f5deb796 HY	67	pud = (pud_t *)get_zeroed_page(GFP_KERNEL);
	68	if (!pud)
	69	goto err;
	70	image->arch.pud = pud;
7f689041	71	set_p4d(p4d, __p4d(__pa(pud) \| _KERNPG_TABLE));
f5deb796	72	}
7f689041	73	pud = pud_offset(p4d, vaddr);
f5deb796 HY	74	if (!pud_present(*pud)) {
	75	pmd = (pmd_t *)get_zeroed_page(GFP_KERNEL);
	76	if (!pmd)
	77	goto err;
	78	image->arch.pmd = pmd;
	79	set_pud(pud, __pud(__pa(pmd) \| _KERNPG_TABLE));
	80	}
	81	pmd = pmd_offset(pud, vaddr);
	82	if (!pmd_present(*pmd)) {
	83	pte = (pte_t *)get_zeroed_page(GFP_KERNEL);
	84	if (!pte)
	85	goto err;
	86	image->arch.pte = pte;
	87	set_pmd(pmd, __pmd(__pa(pte) \| _KERNPG_TABLE));
	88	}
	89	pte = pte_offset_kernel(pmd, vaddr);
bba4ed01	90	set_pte(pte, pfn_pte(paddr >> PAGE_SHIFT, PAGE_KERNEL_EXEC_NOENC));
f5deb796 HY	91	return 0;
	92	err:
	93	free_transition_pgtable(image);
	94	return result;
	95	}
	96
9ebdc79f YL	97	static void alloc_pgt_page(void data)
	98	{
	99	struct kimage image = (struct kimage )data;
	100	struct page *page;
	101	void *p = NULL;
	102
	103	page = kimage_alloc_control_pages(image, 0);
	104	if (page) {
	105	p = page_address(page);
	106	clear_page(p);
	107	}
	108
	109	return p;
	110	}
	111
5234f5eb EB	112	static int init_pgtable(struct kimage *image, unsigned long start_pgtable)
5234f5eb EB	113	{
9ebdc79f YL	114	struct x86_mapping_info info = {
	115	.alloc_pgt_page = alloc_pgt_page,
	116	.context = image,
66aad4fd	117	.page_flag = __PAGE_KERNEL_LARGE_EXEC,
bba4ed01	118	.kernpg_flag = _KERNPG_TABLE_NOENC,
9ebdc79f	119	};
084d1283	120	unsigned long mstart, mend;
8bf27556	121	pgd_t *level4p;
f5deb796	122	int result;
084d1283 YL	123	int i;
084d1283 YL	124
8bf27556	125	level4p = (pgd_t *)__va(start_pgtable);
9ebdc79f	126	clear_page(level4p);
8638100c XP	127
	128	if (direct_gbpages)
	129	info.direct_gbpages = true;
	130
0e691cf8 YL	131	for (i = 0; i < nr_pfn_mapped; i++) {
	132	mstart = pfn_mapped[i].start << PAGE_SHIFT;
	133	mend = pfn_mapped[i].end << PAGE_SHIFT;
	134
	135	result = kernel_ident_mapping_init(&info,
	136	level4p, mstart, mend);
	137	if (result)
	138	return result;
	139	}
084d1283	140
53594547	141	/*
084d1283 YL	142	* segments's mem ranges could be outside 0 ~ max_pfn,
	143	* for example when jump back to original kernel from kexeced kernel.
	144	* or first kernel is booted with user mem map, and second kernel
	145	* could be loaded out of that range.
53594547	146	*/
084d1283 YL	147	for (i = 0; i < image->nr_segments; i++) {
	148	mstart = image->segment[i].mem;
	149	mend = mstart + image->segment[i].memsz;
	150
9ebdc79f YL	151	result = kernel_ident_mapping_init(&info,
9ebdc79f YL	152	level4p, mstart, mend);
084d1283 YL	153
	154	if (result)
	155	return result;
	156	}
	157
f5deb796	158	return init_transition_pgtable(image, level4p);
5234f5eb EB	159	}
	160
	161	static void set_idt(void *newidt, u16 limit)
	162	{
36c4fd23	163	struct desc_ptr curidt;
5234f5eb EB	164
5234f5eb EB	165	/* x86-64 supports unaliged loads & stores */
36c4fd23 EB	166	curidt.size = limit;
36c4fd23 EB	167	curidt.address = (unsigned long)newidt;
5234f5eb EB	168
5234f5eb EB	169	__asm__ __volatile__ (
36c4fd23 EB	170	"lidtq %0\n"
36c4fd23 EB	171	: : "m" (curidt)
5234f5eb EB	172	);
	173	};
	174
	175
	176	static void set_gdt(void *newgdt, u16 limit)
	177	{
36c4fd23	178	struct desc_ptr curgdt;
5234f5eb EB	179
5234f5eb EB	180	/* x86-64 supports unaligned loads & stores */
36c4fd23 EB	181	curgdt.size = limit;
36c4fd23 EB	182	curgdt.address = (unsigned long)newgdt;
5234f5eb EB	183
5234f5eb EB	184	__asm__ __volatile__ (
36c4fd23 EB	185	"lgdtq %0\n"
36c4fd23 EB	186	: : "m" (curgdt)
5234f5eb EB	187	);
	188	};
	189
	190	static void load_segments(void)
	191	{
	192	__asm__ __volatile__ (
36c4fd23 EB	193	"\tmovl %0,%%ds\n"
	194	"\tmovl %0,%%es\n"
	195	"\tmovl %0,%%ss\n"
	196	"\tmovl %0,%%fs\n"
	197	"\tmovl %0,%%gs\n"
2ec5e3a8	198	: : "a" (__KERNEL_DS) : "memory"
5234f5eb	199	);
5234f5eb EB	200	}
5234f5eb EB	201
74ca317c	202	#ifdef CONFIG_KEXEC_FILE
dd5f7260 VG	203	/* Update purgatory as needed after various image segments have been prepared */
	204	static int arch_update_purgatory(struct kimage *image)
	205	{
	206	int ret = 0;
	207
	208	if (!image->file_mode)
	209	return 0;
	210
	211	/* Setup copying of backup region */
	212	if (image->type == KEXEC_TYPE_CRASH) {
40c50c1f TG	213	ret = kexec_purgatory_get_set_symbol(image,
40c50c1f TG	214	"purgatory_backup_dest",
dd5f7260 VG	215	&image->arch.backup_load_addr,
	216	sizeof(image->arch.backup_load_addr), 0);
	217	if (ret)
	218	return ret;
	219
40c50c1f TG	220	ret = kexec_purgatory_get_set_symbol(image,
40c50c1f TG	221	"purgatory_backup_src",
dd5f7260 VG	222	&image->arch.backup_src_start,
	223	sizeof(image->arch.backup_src_start), 0);
	224	if (ret)
	225	return ret;
	226
40c50c1f TG	227	ret = kexec_purgatory_get_set_symbol(image,
40c50c1f TG	228	"purgatory_backup_sz",
dd5f7260 VG	229	&image->arch.backup_src_sz,
	230	sizeof(image->arch.backup_src_sz), 0);
	231	if (ret)
	232	return ret;
	233	}
	234
	235	return ret;
	236	}
74ca317c VG	237	#else /* !CONFIG_KEXEC_FILE */
	238	static inline int arch_update_purgatory(struct kimage *image)
	239	{
	240	return 0;
	241	}
	242	#endif /* CONFIG_KEXEC_FILE */
dd5f7260	243
5234f5eb EB	244	int machine_kexec_prepare(struct kimage *image)
5234f5eb EB	245	{
4bfaaef0	246	unsigned long start_pgtable;
5234f5eb EB	247	int result;
	248
	249	/* Calculate the offsets */
72414d3f	250	start_pgtable = page_to_pfn(image->control_code_page) << PAGE_SHIFT;
5234f5eb EB	251
	252	/* Setup the identity mapped 64bit page table */
	253	result = init_pgtable(image, start_pgtable);
72414d3f	254	if (result)
5234f5eb	255	return result;
5234f5eb	256
dd5f7260 VG	257	/* update purgatory as needed */
	258	result = arch_update_purgatory(image);
	259	if (result)
	260	return result;
	261
5234f5eb EB	262	return 0;
	263	}
	264
	265	void machine_kexec_cleanup(struct kimage *image)
	266	{
f5deb796	267	free_transition_pgtable(image);
5234f5eb EB	268	}
	269
	270	/*
	271	* Do not allocate memory (or fail in any way) in machine_kexec().
	272	* We are past the point of no return, committed to rebooting now.
	273	*/
3ab83521	274	void machine_kexec(struct kimage *image)
5234f5eb	275	{
4bfaaef0 MD	276	unsigned long page_list[PAGES_NR];
4bfaaef0 MD	277	void *control_page;
fee7b0d8	278	int save_ftrace_enabled;
5234f5eb	279
fee7b0d8	280	#ifdef CONFIG_KEXEC_JUMP
6407df5c	281	if (image->preserve_context)
fee7b0d8 HY	282	save_processor_state();
	283	#endif
	284
	285	save_ftrace_enabled = __ftrace_enabled_save();
f43fdad8	286
5234f5eb EB	287	/* Interrupts aren't acceptable while we reboot */
5234f5eb EB	288	local_irq_disable();
17f557e5	289	hw_breakpoint_disable();
5234f5eb	290
fee7b0d8 HY	291	if (image->preserve_context) {
	292	#ifdef CONFIG_X86_IO_APIC
	293	/*
	294	* We need to put APICs in legacy mode so that we can
	295	* get timer interrupts in second kernel. kexec/kdump
	296	* paths already have calls to disable_IO_APIC() in
	297	* one form or other. kexec jump path also need
	298	* one.
	299	*/
	300	disable_IO_APIC();
	301	#endif
	302	}
	303
4bfaaef0	304	control_page = page_address(image->control_code_page) + PAGE_SIZE;
fee7b0d8	305	memcpy(control_page, relocate_kernel, KEXEC_CONTROL_CODE_MAX_SIZE);
4bfaaef0	306
e3ebadd9	307	page_list[PA_CONTROL_PAGE] = virt_to_phys(control_page);
fee7b0d8	308	page_list[VA_CONTROL_PAGE] = (unsigned long)control_page;
4bfaaef0 MD	309	page_list[PA_TABLE_PAGE] =
4bfaaef0 MD	310	(unsigned long)__pa(page_address(image->control_code_page));
5234f5eb	311
fee7b0d8 HY	312	if (image->type == KEXEC_TYPE_DEFAULT)
	313	page_list[PA_SWAP_PAGE] = (page_to_pfn(image->swap_page)
	314	<< PAGE_SHIFT);
	315
fef3a7a1 HY	316	/*
fef3a7a1 HY	317	* The segment registers are funny things, they have both a
2a8a3d5b EB	318	* visible and an invisible part. Whenever the visible part is
	319	* set to a specific selector, the invisible part is loaded
	320	* with from a table in memory. At no other time is the
	321	* descriptor table in memory accessed.
5234f5eb EB	322	*
	323	* I take advantage of this here by force loading the
	324	* segments, before I zap the gdt with an invalid value.
	325	*/
	326	load_segments();
fef3a7a1 HY	327	/*
fef3a7a1 HY	328	* The gdt & idt are now invalid.
5234f5eb EB	329	* If you want to load them you must set up your own idt & gdt.
5234f5eb EB	330	*/
fef3a7a1 HY	331	set_gdt(phys_to_virt(0), 0);
fef3a7a1 HY	332	set_idt(phys_to_virt(0), 0);
4bfaaef0	333
5234f5eb	334	/* now call it */
fee7b0d8 HY	335	image->start = relocate_kernel((unsigned long)image->head,
	336	(unsigned long)page_list,
	337	image->start,
4e237903 TL	338	image->preserve_context,
4e237903 TL	339	sme_active());
fee7b0d8 HY	340
fee7b0d8 HY	341	#ifdef CONFIG_KEXEC_JUMP
6407df5c	342	if (image->preserve_context)
fee7b0d8 HY	343	restore_processor_state();
	344	#endif
	345
	346	__ftrace_enabled_restore(save_ftrace_enabled);
5234f5eb	347	}
2c8c0e6b	348
fd59d231 KO	349	void arch_crash_save_vmcoreinfo(void)
fd59d231 KO	350	{
401721ec	351	VMCOREINFO_NUMBER(phys_base);
65ade2f8	352	VMCOREINFO_SYMBOL(init_top_pgt);
92df5c3e KO	353
	354	#ifdef CONFIG_NUMA
	355	VMCOREINFO_SYMBOL(node_data);
	356	VMCOREINFO_LENGTH(node_data, MAX_NUMNODES);
	357	#endif
b6085a86	358	vmcoreinfo_append_str("KERNELOFFSET=%lx\n",
4545c898	359	kaslr_offset());
401721ec	360	VMCOREINFO_NUMBER(KERNEL_IMAGE_SIZE);
fd59d231 KO	361	}
fd59d231 KO	362
cb105258 VG	363	/* arch-dependent functionality related to kexec file-based syscall */
cb105258 VG	364
74ca317c	365	#ifdef CONFIG_KEXEC_FILE
cb105258 VG	366	int arch_kexec_kernel_image_probe(struct kimage image, void buf,
	367	unsigned long buf_len)
	368	{
	369	int i, ret = -ENOEXEC;
	370	struct kexec_file_ops *fops;
	371
	372	for (i = 0; i < ARRAY_SIZE(kexec_file_loaders); i++) {
	373	fops = kexec_file_loaders[i];
	374	if (!fops \|\| !fops->probe)
	375	continue;
	376
	377	ret = fops->probe(buf, buf_len);
	378	if (!ret) {
	379	image->fops = fops;
	380	return ret;
	381	}
	382	}
	383
	384	return ret;
	385	}
	386
	387	void arch_kexec_kernel_image_load(struct kimage image)
	388	{
dd5f7260 VG	389	vfree(image->arch.elf_headers);
	390	image->arch.elf_headers = NULL;
	391
cb105258 VG	392	if (!image->fops \|\| !image->fops->load)
	393	return ERR_PTR(-ENOEXEC);
	394
	395	return image->fops->load(image, image->kernel_buf,
	396	image->kernel_buf_len, image->initrd_buf,
	397	image->initrd_buf_len, image->cmdline_buf,
	398	image->cmdline_buf_len);
	399	}
	400
	401	int arch_kimage_file_post_load_cleanup(struct kimage *image)
	402	{
	403	if (!image->fops \|\| !image->fops->cleanup)
	404	return 0;
	405
27f48d3e	406	return image->fops->cleanup(image->image_loader_data);
cb105258	407	}
12db5562	408
978e30c9	409	#ifdef CONFIG_KEXEC_VERIFY_SIG
8e7d8381 VG	410	int arch_kexec_kernel_verify_sig(struct kimage image, void kernel,
	411	unsigned long kernel_len)
	412	{
	413	if (!image->fops \|\| !image->fops->verify_sig) {
	414	pr_debug("kernel loader does not support signature verification.");
	415	return -EKEYREJECTED;
	416	}
	417
	418	return image->fops->verify_sig(kernel, kernel_len);
	419	}
978e30c9	420	#endif
8e7d8381	421
12db5562 VG	422	/*
	423	* Apply purgatory relocations.
	424	*
	425	* ehdr: Pointer to elf headers
	426	* sechdrs: Pointer to section headers.
	427	* relsec: section index of SHT_RELA section.
	428	*
	429	* TODO: Some of the code belongs to generic code. Move that in kexec.c.
	430	*/
	431	int arch_kexec_apply_relocations_add(const Elf64_Ehdr *ehdr,
	432	Elf64_Shdr *sechdrs, unsigned int relsec)
	433	{
	434	unsigned int i;
	435	Elf64_Rela *rel;
	436	Elf64_Sym *sym;
	437	void *location;
	438	Elf64_Shdr section, symtabsec;
	439	unsigned long address, sec_base, value;
	440	const char strtab, name, *shstrtab;
	441
	442	/*
	443	* ->sh_offset has been modified to keep the pointer to section
	444	* contents in memory
	445	*/
	446	rel = (void *)sechdrs[relsec].sh_offset;
	447
	448	/* Section to which relocations apply */
	449	section = &sechdrs[sechdrs[relsec].sh_info];
	450
	451	pr_debug("Applying relocate section %u to %u\n", relsec,
	452	sechdrs[relsec].sh_info);
	453
	454	/* Associated symbol table */
	455	symtabsec = &sechdrs[sechdrs[relsec].sh_link];
	456
	457	/* String table */
	458	if (symtabsec->sh_link >= ehdr->e_shnum) {
	459	/* Invalid strtab section number */
	460	pr_err("Invalid string table section index %d\n",
	461	symtabsec->sh_link);
	462	return -ENOEXEC;
	463	}
	464
	465	strtab = (char *)sechdrs[symtabsec->sh_link].sh_offset;
	466
	467	/* section header string table */
	468	shstrtab = (char *)sechdrs[ehdr->e_shstrndx].sh_offset;
	469
	470	for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rel); i++) {
	471
	472	/*
	473	* rel[i].r_offset contains byte offset from beginning
	474	* of section to the storage unit affected.
	475	*
	476	* This is location to update (->sh_offset). This is temporary
	477	* buffer where section is currently loaded. This will finally
	478	* be loaded to a different address later, pointed to by
	479	* ->sh_addr. kexec takes care of moving it
	480	* (kexec_load_segment()).
	481	*/
	482	location = (void *)(section->sh_offset + rel[i].r_offset);
	483
	484	/* Final address of the location */
	485	address = section->sh_addr + rel[i].r_offset;
486
487	/*
488	* rel[i].r_info contains information about symbol table index
489	* w.r.t which relocation must be made and type of relocation
490	* to apply. ELF64_R_SYM() and ELF64_R_TYPE() macros get
491	* these respectively.
492	*/
493	sym = (Elf64_Sym *)symtabsec->sh_offset +
494	ELF64_R_SYM(rel[i].r_info);
495
496	if (sym->st_name)
497	name = strtab + sym->st_name;
498	else
499	name = shstrtab + sechdrs[sym->st_shndx].sh_name;
500
501	pr_debug("Symbol: %s info: %02x shndx: %02x value=%llx size: %llx\n",
502	name, sym->st_info, sym->st_shndx, sym->st_value,
503	sym->st_size);
504
505	if (sym->st_shndx == SHN_UNDEF) {
506	pr_err("Undefined symbol: %s\n", name);
507	return -ENOEXEC;
508	}
509
510	if (sym->st_shndx == SHN_COMMON) {
511	pr_err("symbol '%s' in common section\n", name);
512	return -ENOEXEC;
513	}
514
515	if (sym->st_shndx == SHN_ABS)
516	sec_base = 0;
517	else if (sym->st_shndx >= ehdr->e_shnum) {
518	pr_err("Invalid section %d for symbol %s\n",
519	sym->st_shndx, name);
520	return -ENOEXEC;
521	} else
522	sec_base = sechdrs[sym->st_shndx].sh_addr;
523
524	value = sym->st_value;
525	value += sec_base;
526	value += rel[i].r_addend;
527
528	switch (ELF64_R_TYPE(rel[i].r_info)) {
529	case R_X86_64_NONE:
530	break;
531	case R_X86_64_64:
532	(u64 )location = value;
533	break;
534	case R_X86_64_32:
535	(u32 )location = value;
536	if (value != (u32 )location)
537	goto overflow;
538	break;
539	case R_X86_64_32S:
540	(s32 )location = value;
541	if ((s64)value != (s32 )location)
542	goto overflow;
543	break;
544	case R_X86_64_PC32:
545	value -= (u64)address;
546	(u32 )location = value;
547	break;
548	default:
549	pr_err("Unknown rela relocation: %llu\n",
550	ELF64_R_TYPE(rel[i].r_info));
551	return -ENOEXEC;
552	}
553	}
554	return 0;
555
556	overflow:
557	pr_err("Overflow in relocation type %d value 0x%lx\n",
558	(int)ELF64_R_TYPE(rel[i].r_info), value);
559	return -ENOEXEC;
560	}
74ca317c	561	#endif /* CONFIG_KEXEC_FILE */
1e5768ae XP	562
	563	static int
	564	kexec_mark_range(unsigned long start, unsigned long end, bool protect)
	565	{
	566	struct page *page;
	567	unsigned int nr_pages;
	568
	569	/*
	570	* For physical range: [start, end]. We must skip the unassigned
	571	* crashk resource with zero-valued "end" member.
	572	*/
	573	if (!end \|\| start > end)
	574	return 0;
	575
	576	page = pfn_to_page(start >> PAGE_SHIFT);
	577	nr_pages = (end >> PAGE_SHIFT) - (start >> PAGE_SHIFT) + 1;
	578	if (protect)
	579	return set_pages_ro(page, nr_pages);
	580	else
	581	return set_pages_rw(page, nr_pages);
	582	}
	583
	584	static void kexec_mark_crashkres(bool protect)
	585	{
	586	unsigned long control;
	587
	588	kexec_mark_range(crashk_low_res.start, crashk_low_res.end, protect);
	589
	590	/* Don't touch the control code page used in crash_kexec().*/
	591	control = PFN_PHYS(page_to_pfn(kexec_crash_image->control_code_page));
	592	/* Control code page is located in the 2nd page. */
	593	kexec_mark_range(crashk_res.start, control + PAGE_SIZE - 1, protect);
	594	control += KEXEC_CONTROL_PAGE_SIZE;
	595	kexec_mark_range(control, crashk_res.end, protect);
	596	}
	597
	598	void arch_kexec_protect_crashkres(void)
	599	{
	600	kexec_mark_crashkres(true);
	601	}
	602
	603	void arch_kexec_unprotect_crashkres(void)
	604	{
	605	kexec_mark_crashkres(false);
	606	}
bba4ed01 TL	607
	608	int arch_kexec_post_alloc_pages(void *vaddr, unsigned int pages, gfp_t gfp)
	609	{
	610	/*
	611	* If SME is active we need to be sure that kexec pages are
	612	* not encrypted because when we boot to the new kernel the
	613	* pages won't be accessed encrypted (initially).
	614	*/
	615	return set_memory_decrypted((unsigned long)vaddr, pages);
	616	}
	617
	618	void arch_kexec_pre_free_pages(void *vaddr, unsigned int pages)
	619	{
	620	/*
	621	* If SME is active we need to reset the pages back to being
	622	* an encrypted mapping before freeing them.
	623	*/
	624	set_memory_encrypted((unsigned long)vaddr, pages);
	625	}