[linux-2.6-block.git] / arch / x86 / boot / compressed / misc.c

// SPDX-License-Identifier: GPL-2.0
/*
 * misc.c
 *
 * This is a collection of several routines used to extract the kernel
 * which includes KASLR relocation, decompression, ELF parsing, and
 * relocation processing. Additionally included are the screen and serial
 * output functions and related debugging support functions.
 *
 * malloc by Hannu Savolainen 1993 and Matthias Urlichs 1994
 * puts by Nick Holloway 1993, better puts by Martin Mares 1995
 * High loaded stuff by Hans Lermen & Werner Almesberger, Feb. 1996
 */

#include "misc.h"
#include "error.h"
#include "../string.h"
#include "../voffset.h"
#include <asm/bootparam_utils.h>

/*
 * WARNING!!
 * This code is compiled with -fPIC and it is relocated dynamically at
 * run time, but no relocation processing is performed. This means that
 * it is not safe to place pointers in static structures.
 */

/* Macros used by the included decompressor code below. */
#define STATIC		static
/* Define an externally visible malloc()/free(). */
#define MALLOC_VISIBLE
#include <linux/decompress/mm.h>

/*
 * Provide definitions of memzero and memmove as some of the decompressors will
 * try to define their own functions if these are not defined as macros.
 */
#define memzero(s, n)	memset((s), 0, (n))
#ifndef memmove
#define memmove		memmove
/* Functions used by the included decompressor code below. */
void *memmove(void *dest, const void *src, size_t n);
#endif

/*
 * This is set up by the setup-routine at boot-time
 */
struct boot_params *boot_params_ptr;

struct port_io_ops pio_ops;

memptr free_mem_ptr;
memptr free_mem_end_ptr;
int spurious_nmi_count;

static char *vidmem;
static int vidport;

/* These might be accessed before .bss is cleared, so use .data instead. */
static int lines __section(".data");
static int cols __section(".data");

#ifdef CONFIG_KERNEL_GZIP
#include "../../../../lib/decompress_inflate.c"
#endif

#ifdef CONFIG_KERNEL_BZIP2
#include "../../../../lib/decompress_bunzip2.c"
#endif

#ifdef CONFIG_KERNEL_LZMA
#include "../../../../lib/decompress_unlzma.c"
#endif

#ifdef CONFIG_KERNEL_XZ
#include "../../../../lib/decompress_unxz.c"
#endif

#ifdef CONFIG_KERNEL_LZO
#include "../../../../lib/decompress_unlzo.c"
#endif

#ifdef CONFIG_KERNEL_LZ4
#include "../../../../lib/decompress_unlz4.c"
#endif

#ifdef CONFIG_KERNEL_ZSTD
#include "../../../../lib/decompress_unzstd.c"
#endif
/*
 * NOTE: When adding a new decompressor, please update the analysis in
 * ../header.S.
 */

static void scroll(void)
{
	int i;

	memmove(vidmem, vidmem + cols * 2, (lines - 1) * cols * 2);
	for (i = (lines - 1) * cols * 2; i < lines * cols * 2; i += 2)
		vidmem[i] = ' ';
}

#define XMTRDY          0x20

#define TXR             0       /*  Transmit register (WRITE) */
#define LSR             5       /*  Line Status               */
static void serial_putchar(int ch)
{
	unsigned timeout = 0xffff;

	while ((inb(early_serial_base + LSR) & XMTRDY) == 0 && --timeout)
		cpu_relax();

	outb(ch, early_serial_base + TXR);
}

void __putstr(const char *s)
{
	int x, y, pos;
	char c;

	if (early_serial_base) {
		const char *str = s;
		while (*str) {
			if (*str == '\n')
				serial_putchar('\r');
			serial_putchar(*str++);
		}
	}

	if (lines == 0 || cols == 0)
		return;

	x = boot_params_ptr->screen_info.orig_x;
	y = boot_params_ptr->screen_info.orig_y;

	while ((c = *s++) != '\0') {
		if (c == '\n') {
			x = 0;
			if (++y >= lines) {
				scroll();
				y--;
			}
		} else {
			vidmem[(x + cols * y) * 2] = c;
			if (++x >= cols) {
				x = 0;
				if (++y >= lines) {
					scroll();
					y--;
				}
			}
		}
	}

	boot_params_ptr->screen_info.orig_x = x;
	boot_params_ptr->screen_info.orig_y = y;

	pos = (x + cols * y) * 2;	/* Update cursor position */
	outb(14, vidport);
	outb(0xff & (pos >> 9), vidport+1);
	outb(15, vidport);
	outb(0xff & (pos >> 1), vidport+1);
}

static noinline void __putnum(unsigned long value, unsigned int base,
			      int mindig)
{
	char buf[8*sizeof(value)+1];
	char *p;

	p = buf + sizeof(buf);
	*--p = '\0';

	while (mindig-- > 0 || value) {
		unsigned char digit = value % base;
		digit += (digit >= 10) ? ('a'-10) : '0';
		*--p = digit;

		value /= base;
	}

	__putstr(p);
}

void __puthex(unsigned long value)
{
	__putnum(value, 16, sizeof(value)*2);
}

void __putdec(unsigned long value)
{
	__putnum(value, 10, 1);
}

#ifdef CONFIG_X86_NEED_RELOCS
static void handle_relocations(void *output, unsigned long output_len,
			       unsigned long virt_addr)
{
	int *reloc;
	unsigned long delta, map, ptr;
	unsigned long min_addr = (unsigned long)output;
	unsigned long max_addr = min_addr + (VO___bss_start - VO__text);

	/*
	 * Calculate the delta between where vmlinux was linked to load
	 * and where it was actually loaded.
	 */
	delta = min_addr - LOAD_PHYSICAL_ADDR;

	/*
	 * The kernel contains a table of relocation addresses. Those
	 * addresses have the final load address of the kernel in virtual
	 * memory. We are currently working in the self map. So we need to
	 * create an adjustment for kernel memory addresses to the self map.
	 * This will involve subtracting out the base address of the kernel.
	 */
	map = delta - __START_KERNEL_map;

	/*
	 * 32-bit always performs relocations. 64-bit relocations are only
	 * needed if KASLR has chosen a different starting address offset
	 * from __START_KERNEL_map.
	 */
	if (IS_ENABLED(CONFIG_X86_64))
		delta = virt_addr - LOAD_PHYSICAL_ADDR;

	if (!delta) {
		debug_putstr("No relocation needed... ");
		return;
	}
	debug_putstr("Performing relocations... ");

	/*
	 * Process relocations: 32 bit relocations first then 64 bit after.
	 * Two sets of binary relocations are added to the end of the kernel
	 * before compression. Each relocation table entry is the kernel
	 * address of the location which needs to be updated stored as a
	 * 32-bit value which is sign extended to 64 bits.
	 *
	 * Format is:
	 *
	 * kernel bits...
	 * 0 - zero terminator for 64 bit relocations
	 * 64 bit relocation repeated
	 * 0 - zero terminator for 32 bit relocations
	 * 32 bit relocation repeated
	 *
	 * So we work backwards from the end of the decompressed image.
	 */
	for (reloc = output + output_len - sizeof(*reloc); *reloc; reloc--) {
		long extended = *reloc;
		extended += map;

		ptr = (unsigned long)extended;
		if (ptr < min_addr || ptr > max_addr)
			error("32-bit relocation outside of kernel!\n");

		*(uint32_t *)ptr += delta;
	}
#ifdef CONFIG_X86_64
	for (reloc--; *reloc; reloc--) {
		long extended = *reloc;
		extended += map;

		ptr = (unsigned long)extended;
		if (ptr < min_addr || ptr > max_addr)
			error("64-bit relocation outside of kernel!\n");

		*(uint64_t *)ptr += delta;
	}
#endif
}
#else
static inline void handle_relocations(void *output, unsigned long output_len,
				      unsigned long virt_addr)
{ }
#endif

static size_t parse_elf(void *output)
{
#ifdef CONFIG_X86_64
	Elf64_Ehdr ehdr;
	Elf64_Phdr *phdrs, *phdr;
#else
	Elf32_Ehdr ehdr;
	Elf32_Phdr *phdrs, *phdr;
#endif
	void *dest;
	int i;

	memcpy(&ehdr, output, sizeof(ehdr));
	if (ehdr.e_ident[EI_MAG0] != ELFMAG0 ||
	   ehdr.e_ident[EI_MAG1] != ELFMAG1 ||
	   ehdr.e_ident[EI_MAG2] != ELFMAG2 ||
	   ehdr.e_ident[EI_MAG3] != ELFMAG3)
		error("Kernel is not a valid ELF file");

	debug_putstr("Parsing ELF... ");

	phdrs = malloc(sizeof(*phdrs) * ehdr.e_phnum);
	if (!phdrs)
		error("Failed to allocate space for phdrs");

	memcpy(phdrs, output + ehdr.e_phoff, sizeof(*phdrs) * ehdr.e_phnum);

	for (i = 0; i < ehdr.e_phnum; i++) {
		phdr = &phdrs[i];

		switch (phdr->p_type) {
		case PT_LOAD:
#ifdef CONFIG_X86_64
			if ((phdr->p_align % 0x200000) != 0)
				error("Alignment of LOAD segment isn't multiple of 2MB");
#endif
#ifdef CONFIG_RELOCATABLE
			dest = output;
			dest += (phdr->p_paddr - LOAD_PHYSICAL_ADDR);
#else
			dest = (void *)(phdr->p_paddr);
#endif
			memmove(dest, output + phdr->p_offset, phdr->p_filesz);
			break;
		default: /* Ignore other PT_* */ break;
		}
	}

	free(phdrs);

	return ehdr.e_entry - LOAD_PHYSICAL_ADDR;
}

const unsigned long kernel_text_size = VO___start_rodata - VO__text;
const unsigned long kernel_total_size = VO__end - VO__text;

static u8 boot_heap[BOOT_HEAP_SIZE] __aligned(4);

extern unsigned char input_data[];
extern unsigned int input_len, output_len;

unsigned long decompress_kernel(unsigned char *outbuf, unsigned long virt_addr,
				void (*error)(char *x))
{
	unsigned long entry;

	if (!free_mem_ptr) {
		free_mem_ptr     = (unsigned long)boot_heap;
		free_mem_end_ptr = (unsigned long)boot_heap + sizeof(boot_heap);
	}

	if (__decompress(input_data, input_len, NULL, NULL, outbuf, output_len,
			 NULL, error) < 0)
		return ULONG_MAX;

	entry = parse_elf(outbuf);
	handle_relocations(outbuf, output_len, virt_addr);

	return entry;
}

/*
 * Set the memory encryption xloadflag based on the mem_encrypt= command line
 * parameter, if provided.
 */
static void parse_mem_encrypt(struct setup_header *hdr)
{
	int on = cmdline_find_option_bool("mem_encrypt=on");
	int off = cmdline_find_option_bool("mem_encrypt=off");

	if (on > off)
		hdr->xloadflags |= XLF_MEM_ENCRYPTION;
}

static void early_sev_detect(void)
{
	/*
	 * Accessing video memory causes guest termination because
	 * the boot stage2 #VC handler of SEV-ES/SNP guests does not
	 * support MMIO handling and kexec -c adds screen_info to the
	 * boot parameters passed to the kexec kernel, which causes
	 * console output to be dumped to both video and serial.
	 */
	if (sev_status & MSR_AMD64_SEV_ES_ENABLED)
		lines = cols = 0;
}

/*
 * The compressed kernel image (ZO), has been moved so that its position
 * is against the end of the buffer used to hold the uncompressed kernel
 * image (VO) and the execution environment (.bss, .brk), which makes sure
 * there is room to do the in-place decompression. (See header.S for the
 * calculations.)
 *
 *                             |-----compressed kernel image------|
 *                             V                                  V
 * 0                       extract_offset                      +INIT_SIZE
 * |-----------|---------------|-------------------------|--------|
 *             |               |                         |        |
 *           VO__text      startup_32 of ZO          VO__end    ZO__end
 *             ^                                         ^
 *             |-------uncompressed kernel image---------|
 *
 */
asmlinkage __visible void *extract_kernel(void *rmode, unsigned char *output)
{
	unsigned long virt_addr = LOAD_PHYSICAL_ADDR;
	memptr heap = (memptr)boot_heap;
	unsigned long needed_size;
	size_t entry_offset;

	/* Retain x86 boot parameters pointer passed from startup_32/64. */
	boot_params_ptr = rmode;

	/* Clear flags intended for solely in-kernel use. */
	boot_params_ptr->hdr.loadflags &= ~KASLR_FLAG;

	parse_mem_encrypt(&boot_params_ptr->hdr);

	sanitize_boot_params(boot_params_ptr);

	if (boot_params_ptr->screen_info.orig_video_mode == 7) {
		vidmem = (char *) 0xb0000;
		vidport = 0x3b4;
	} else {
		vidmem = (char *) 0xb8000;
		vidport = 0x3d4;
	}

	lines = boot_params_ptr->screen_info.orig_video_lines;
	cols = boot_params_ptr->screen_info.orig_video_cols;

	init_default_io_ops();

	/*
	 * Detect TDX guest environment.
	 *
	 * It has to be done before console_init() in order to use
	 * paravirtualized port I/O operations if needed.
	 */
	early_tdx_detect();

	early_sev_detect();

	console_init();

	/*
	 * Save RSDP address for later use. Have this after console_init()
	 * so that early debugging output from the RSDP parsing code can be
	 * collected.
	 */
	boot_params_ptr->acpi_rsdp_addr = get_rsdp_addr();

	debug_putstr("early console in extract_kernel\n");

	free_mem_ptr     = heap;	/* Heap */
	free_mem_end_ptr = heap + BOOT_HEAP_SIZE;

	/*
	 * The memory hole needed for the kernel is the larger of either
	 * the entire decompressed kernel plus relocation table, or the
	 * entire decompressed kernel plus .bss and .brk sections.
	 *
	 * On X86_64, the memory is mapped with PMD pages. Round the
	 * size up so that the full extent of PMD pages mapped is
	 * included in the check against the valid memory table
	 * entries. This ensures the full mapped area is usable RAM
	 * and doesn't include any reserved areas.
	 */
	needed_size = max_t(unsigned long, output_len, kernel_total_size);
#ifdef CONFIG_X86_64
	needed_size = ALIGN(needed_size, MIN_KERNEL_ALIGN);
#endif

	/* Report initial kernel position details. */
	debug_putaddr(input_data);
	debug_putaddr(input_len);
	debug_putaddr(output);
	debug_putaddr(output_len);
	debug_putaddr(kernel_total_size);
	debug_putaddr(needed_size);

#ifdef CONFIG_X86_64
	/* Report address of 32-bit trampoline */
	debug_putaddr(trampoline_32bit);
#endif

	choose_random_location((unsigned long)input_data, input_len,
				(unsigned long *)&output,
				needed_size,
				&virt_addr);

	/* Validate memory location choices. */
	if ((unsigned long)output & (MIN_KERNEL_ALIGN - 1))
		error("Destination physical address inappropriately aligned");
	if (virt_addr & (MIN_KERNEL_ALIGN - 1))
		error("Destination virtual address inappropriately aligned");
#ifdef CONFIG_X86_64
	if (heap > 0x3fffffffffffUL)
		error("Destination address too large");
	if (virt_addr + needed_size > KERNEL_IMAGE_SIZE)
		error("Destination virtual address is beyond the kernel mapping area");
#else
	if (heap > ((-__PAGE_OFFSET-(128<<20)-1) & 0x7fffffff))
		error("Destination address too large");
#endif
#ifndef CONFIG_RELOCATABLE
	if (virt_addr != LOAD_PHYSICAL_ADDR)
		error("Destination virtual address changed when not relocatable");
#endif

	debug_putstr("\nDecompressing Linux... ");

	if (init_unaccepted_memory()) {
		debug_putstr("Accepting memory... ");
		accept_memory(__pa(output), needed_size);
	}

	entry_offset = decompress_kernel(output, virt_addr, error);

	debug_putstr("done.\nBooting the kernel (entry_offset: 0x");
	debug_puthex(entry_offset);
	debug_putstr(").\n");

	/* Disable exception handling before booting the kernel */
	cleanup_exception_handling();

	if (spurious_nmi_count) {
		error_putstr("Spurious early NMIs ignored: ");
		error_putdec(spurious_nmi_count);
		error_putstr("\n");
	}

	return output + entry_offset;
}
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
1da177e4 LT	2	/*
1da177e4 LT	3	* misc.c
818a08f8	4	*
c0402881 KC	5	* This is a collection of several routines used to extract the kernel
	6	* which includes KASLR relocation, decompression, ELF parsing, and
	7	* relocation processing. Additionally included are the screen and serial
	8	* output functions and related debugging support functions.
1da177e4 LT	9	*
	10	* malloc by Hannu Savolainen 1993 and Matthias Urlichs 1994
	11	* puts by Nick Holloway 1993, better puts by Martin Mares 1995
	12	* High loaded stuff by Hans Lermen & Werner Almesberger, Feb. 1996
	13	*/
	14
8fee13a4	15	#include "misc.h"
dc425a6e	16	#include "error.h"
820e8fec	17	#include "../string.h"
67b66625	18	#include "../voffset.h"
8c5477e8	19	#include <asm/bootparam_utils.h>
968de4f0	20
968de4f0	21	/*
4252db10 BH	22	* WARNING!!
	23	* This code is compiled with -fPIC and it is relocated dynamically at
	24	* run time, but no relocation processing is performed. This means that
	25	* it is not safe to place pointers in static structures.
968de4f0	26	*/
1da177e4	27
1f208de3	28	/* Macros used by the included decompressor code below. */
1180e01d	29	#define STATIC static
33f98a97 KC	30	/* Define an externally visible malloc()/free(). */
	31	#define MALLOC_VISIBLE
	32	#include <linux/decompress/mm.h>
1da177e4	33
04999550	34	/*
394b19d6 AS	35	* Provide definitions of memzero and memmove as some of the decompressors will
394b19d6 AS	36	* try to define their own functions if these are not defined as macros.
04999550	37	*/
1180e01d	38	#define memzero(s, n) memset((s), 0, (n))
938a000e	39	#ifndef memmove
81b785f3	40	#define memmove memmove
1f208de3	41	/* Functions used by the included decompressor code below. */
81b785f3	42	void memmove(void dest, const void *src, size_t n);
938a000e	43	#endif
fd77c7ca	44
1da177e4 LT	45	/*
	46	* This is set up by the setup-routine at boot-time
	47	*/
d55d5bc5	48	struct boot_params *boot_params_ptr;
1da177e4	49
eb4ea1ae KS	50	struct port_io_ops pio_ops;
eb4ea1ae KS	51
82fa9637 KC	52	memptr free_mem_ptr;
82fa9637 KC	53	memptr free_mem_end_ptr;
ac456ca0	54	int spurious_nmi_count;
1da177e4	55
03056c88	56	static char *vidmem;
1da177e4	57	static int vidport;
6044d159 MR	58
	59	/* These might be accessed before .bss is cleared, so use .data instead. */
	60	static int lines __section(".data");
	61	static int cols __section(".data");
1da177e4	62
ae03c499 AK	63	#ifdef CONFIG_KERNEL_GZIP
	64	#include "../../../../lib/decompress_inflate.c"
	65	#endif
	66
	67	#ifdef CONFIG_KERNEL_BZIP2
	68	#include "../../../../lib/decompress_bunzip2.c"
	69	#endif
	70
	71	#ifdef CONFIG_KERNEL_LZMA
	72	#include "../../../../lib/decompress_unlzma.c"
	73	#endif
1da177e4	74
30314804 LC	75	#ifdef CONFIG_KERNEL_XZ
	76	#include "../../../../lib/decompress_unxz.c"
	77	#endif
	78
13510997 AT	79	#ifdef CONFIG_KERNEL_LZO
	80	#include "../../../../lib/decompress_unlzo.c"
	81	#endif
	82
f9b493ac KL	83	#ifdef CONFIG_KERNEL_LZ4
	84	#include "../../../../lib/decompress_unlz4.c"
	85	#endif
fb46d057 NT	86
	87	#ifdef CONFIG_KERNEL_ZSTD
	88	#include "../../../../lib/decompress_unzstd.c"
	89	#endif
4252db10 BH	90	/*
	91	* NOTE: When adding a new decompressor, please update the analysis in
	92	* ../header.S.
	93	*/
f9b493ac	94
1da177e4 LT	95	static void scroll(void)
	96	{
	97	int i;
	98
81b785f3	99	memmove(vidmem, vidmem + cols * 2, (lines - 1) * cols * 2);
fd77c7ca	100	for (i = (lines - 1) * cols * 2; i < lines * cols * 2; i += 2)
1da177e4 LT	101	vidmem[i] = ' ';
	102	}
	103
8fee13a4 YL	104	#define XMTRDY 0x20
	105
	106	#define TXR 0 /* Transmit register (WRITE) */
	107	#define LSR 5 /* Line Status */
	108	static void serial_putchar(int ch)
	109	{
	110	unsigned timeout = 0xffff;
	111
	112	while ((inb(early_serial_base + LSR) & XMTRDY) == 0 && --timeout)
	113	cpu_relax();
	114
	115	outb(ch, early_serial_base + TXR);
	116	}
	117
7aac3015	118	void __putstr(const char *s)
1da177e4	119	{
fd77c7ca	120	int x, y, pos;
1da177e4 LT	121	char c;
1da177e4 LT	122
8fee13a4 YL	123	if (early_serial_base) {
	124	const char *str = s;
	125	while (*str) {
	126	if (*str == '\n')
	127	serial_putchar('\r');
	128	serial_putchar(*str++);
	129	}
	130	}
6bcb13b3	131
fb1cc2f9	132	if (lines == 0 \|\| cols == 0)
a24e7851 RR	133	return;
a24e7851 RR	134
d55d5bc5 AB	135	x = boot_params_ptr->screen_info.orig_x;
d55d5bc5 AB	136	y = boot_params_ptr->screen_info.orig_y;
1da177e4	137
fd77c7ca PC	138	while ((c = *s++) != '\0') {
fd77c7ca PC	139	if (c == '\n') {
1da177e4	140	x = 0;
fd77c7ca	141	if (++y >= lines) {
1da177e4 LT	142	scroll();
	143	y--;
	144	}
	145	} else {
020878ac	146	vidmem[(x + cols * y) * 2] = c;
fd77c7ca	147	if (++x >= cols) {
1da177e4	148	x = 0;
fd77c7ca	149	if (++y >= lines) {
1da177e4 LT	150	scroll();
	151	y--;
	152	}
	153	}
	154	}
	155	}
	156
d55d5bc5 AB	157	boot_params_ptr->screen_info.orig_x = x;
d55d5bc5 AB	158	boot_params_ptr->screen_info.orig_y = y;
1da177e4 LT	159
1da177e4 LT	160	pos = (x + cols * y) * 2; /* Update cursor position */
b02aae9c RH	161	outb(14, vidport);
	162	outb(0xff & (pos >> 9), vidport+1);
	163	outb(15, vidport);
	164	outb(0xff & (pos >> 1), vidport+1);
1da177e4 LT	165	}
1da177e4 LT	166
9ba8ec8e PA	167	static noinline void __putnum(unsigned long value, unsigned int base,
9ba8ec8e PA	168	int mindig)
79063a7c	169	{
9ba8ec8e PA	170	char buf[8*sizeof(value)+1];
9ba8ec8e PA	171	char *p;
79063a7c	172
9ba8ec8e PA	173	p = buf + sizeof(buf);
9ba8ec8e PA	174	*--p = '\0';
79063a7c	175
9ba8ec8e PA	176	while (mindig-- > 0 \|\| value) {
	177	unsigned char digit = value % base;
	178	digit += (digit >= 10) ? ('a'-10) : '0';
	179	*--p = digit;
79063a7c	180
9ba8ec8e	181	value /= base;
79063a7c	182	}
9ba8ec8e PA	183
	184	__putstr(p);
	185	}
	186
	187	void __puthex(unsigned long value)
	188	{
	189	__putnum(value, 16, sizeof(value)*2);
	190	}
	191
	192	void __putdec(unsigned long value)
	193	{
	194	__putnum(value, 10, 1);
79063a7c KC	195	}
79063a7c KC	196
a554e740	197	#ifdef CONFIG_X86_NEED_RELOCS
8391c73c BH	198	static void handle_relocations(void *output, unsigned long output_len,
8391c73c BH	199	unsigned long virt_addr)
a0215061 KC	200	{
	201	int *reloc;
	202	unsigned long delta, map, ptr;
	203	unsigned long min_addr = (unsigned long)output;
4abf061b	204	unsigned long max_addr = min_addr + (VO___bss_start - VO__text);
a0215061 KC	205
	206	/*
	207	* Calculate the delta between where vmlinux was linked to load
	208	* and where it was actually loaded.
	209	*/
	210	delta = min_addr - LOAD_PHYSICAL_ADDR;
a0215061 KC	211
	212	/*
	213	* The kernel contains a table of relocation addresses. Those
	214	* addresses have the final load address of the kernel in virtual
	215	* memory. We are currently working in the self map. So we need to
	216	* create an adjustment for kernel memory addresses to the self map.
	217	* This will involve subtracting out the base address of the kernel.
	218	*/
	219	map = delta - __START_KERNEL_map;
	220
8391c73c BH	221	/*
	222	* 32-bit always performs relocations. 64-bit relocations are only
	223	* needed if KASLR has chosen a different starting address offset
	224	* from __START_KERNEL_map.
	225	*/
	226	if (IS_ENABLED(CONFIG_X86_64))
	227	delta = virt_addr - LOAD_PHYSICAL_ADDR;
	228
	229	if (!delta) {
	230	debug_putstr("No relocation needed... ");
	231	return;
	232	}
	233	debug_putstr("Performing relocations... ");
	234
a0215061 KC	235	/*
a0215061 KC	236	* Process relocations: 32 bit relocations first then 64 bit after.
a8327be7	237	* Two sets of binary relocations are added to the end of the kernel
a0215061 KC	238	* before compression. Each relocation table entry is the kernel
	239	* address of the location which needs to be updated stored as a
	240	* 32-bit value which is sign extended to 64 bits.
	241	*
	242	* Format is:
	243	*
	244	* kernel bits...
	245	* 0 - zero terminator for 64 bit relocations
	246	* 64 bit relocation repeated
	247	* 0 - zero terminator for 32 bit relocations
	248	* 32 bit relocation repeated
	249	*
	250	* So we work backwards from the end of the decompressed image.
	251	*/
	252	for (reloc = output + output_len - sizeof(reloc); reloc; reloc--) {
6f9af75f	253	long extended = *reloc;
a0215061 KC	254	extended += map;
	255
	256	ptr = (unsigned long)extended;
	257	if (ptr < min_addr \|\| ptr > max_addr)
	258	error("32-bit relocation outside of kernel!\n");
	259
	260	(uint32_t )ptr += delta;
	261	}
	262	#ifdef CONFIG_X86_64
	263	for (reloc--; *reloc; reloc--) {
	264	long extended = *reloc;
	265	extended += map;
	266
	267	ptr = (unsigned long)extended;
	268	if (ptr < min_addr \|\| ptr > max_addr)
	269	error("64-bit relocation outside of kernel!\n");
	270
	271	(uint64_t )ptr += delta;
	272	}
	273	#endif
	274	}
	275	#else
8391c73c BH	276	static inline void handle_relocations(void *output, unsigned long output_len,
8391c73c BH	277	unsigned long virt_addr)
a0215061 KC	278	{ }
	279	#endif
	280
7734a0f3	281	static size_t parse_elf(void *output)
099e1377 IC	282	{
	283	#ifdef CONFIG_X86_64
	284	Elf64_Ehdr ehdr;
	285	Elf64_Phdr phdrs, phdr;
	286	#else
	287	Elf32_Ehdr ehdr;
	288	Elf32_Phdr phdrs, phdr;
	289	#endif
	290	void *dest;
	291	int i;
	292
	293	memcpy(&ehdr, output, sizeof(ehdr));
fd77c7ca	294	if (ehdr.e_ident[EI_MAG0] != ELFMAG0 \|\|
099e1377 IC	295	ehdr.e_ident[EI_MAG1] != ELFMAG1 \|\|
099e1377 IC	296	ehdr.e_ident[EI_MAG2] != ELFMAG2 \|\|
7734a0f3	297	ehdr.e_ident[EI_MAG3] != ELFMAG3)
099e1377	298	error("Kernel is not a valid ELF file");
099e1377	299
e605a425	300	debug_putstr("Parsing ELF... ");
099e1377 IC	301
	302	phdrs = malloc(sizeof(phdrs) ehdr.e_phnum);
	303	if (!phdrs)
	304	error("Failed to allocate space for phdrs");
	305
	306	memcpy(phdrs, output + ehdr.e_phoff, sizeof(phdrs) ehdr.e_phnum);
	307
fd77c7ca	308	for (i = 0; i < ehdr.e_phnum; i++) {
099e1377 IC	309	phdr = &phdrs[i];
	310
	311	switch (phdr->p_type) {
	312	case PT_LOAD:
c55b8550 L	313	#ifdef CONFIG_X86_64
	314	if ((phdr->p_align % 0x200000) != 0)
	315	error("Alignment of LOAD segment isn't multiple of 2MB");
	316	#endif
099e1377 IC	317	#ifdef CONFIG_RELOCATABLE
	318	dest = output;
	319	dest += (phdr->p_paddr - LOAD_PHYSICAL_ADDR);
	320	#else
fd77c7ca	321	dest = (void *)(phdr->p_paddr);
099e1377	322	#endif
81b785f3	323	memmove(dest, output + phdr->p_offset, phdr->p_filesz);
099e1377 IC	324	break;
	325	default: /* Ignore other PT_* */ break;
	326	}
	327	}
5067cf53 JJ	328
5067cf53 JJ	329	free(phdrs);
7734a0f3 AL	330
7734a0f3 AL	331	return ehdr.e_entry - LOAD_PHYSICAL_ADDR;
099e1377 IC	332	}
099e1377 IC	333
9c554610	334	const unsigned long kernel_text_size = VO___start_rodata - VO__text;
83381519 AB	335	const unsigned long kernel_total_size = VO__end - VO__text;
83381519 AB	336
24388292 AB	337	static u8 boot_heap[BOOT_HEAP_SIZE] __aligned(4);
	338
	339	extern unsigned char input_data[];
	340	extern unsigned int input_len, output_len;
	341
83381519 AB	342	unsigned long decompress_kernel(unsigned char *outbuf, unsigned long virt_addr,
	343	void (error)(char x))
	344	{
	345	unsigned long entry;
	346
	347	if (!free_mem_ptr) {
	348	free_mem_ptr = (unsigned long)boot_heap;
	349	free_mem_end_ptr = (unsigned long)boot_heap + sizeof(boot_heap);
	350	}
	351
	352	if (__decompress(input_data, input_len, NULL, NULL, outbuf, output_len,
	353	NULL, error) < 0)
	354	return ULONG_MAX;
	355
	356	entry = parse_elf(outbuf);
	357	handle_relocations(outbuf, output_len, virt_addr);
	358
	359	return entry;
	360	}
	361
cd0d9d92 AB	362	/*
	363	* Set the memory encryption xloadflag based on the mem_encrypt= command line
	364	* parameter, if provided.
	365	*/
	366	static void parse_mem_encrypt(struct setup_header *hdr)
	367	{
	368	int on = cmdline_find_option_bool("mem_encrypt=on");
	369	int off = cmdline_find_option_bool("mem_encrypt=off");
	370
	371	if (on > off)
	372	hdr->xloadflags \|= XLF_MEM_ENCRYPTION;
	373	}
	374
f30470c1 AK	375	static void early_sev_detect(void)
	376	{
	377	/*
	378	* Accessing video memory causes guest termination because
	379	* the boot stage2 #VC handler of SEV-ES/SNP guests does not
	380	* support MMIO handling and kexec -c adds screen_info to the
	381	* boot parameters passed to the kexec kernel, which causes
	382	* console output to be dumped to both video and serial.
	383	*/
	384	if (sev_status & MSR_AMD64_SEV_ES_ENABLED)
	385	lines = cols = 0;
	386	}
	387
974f221c YL	388	/*
	389	* The compressed kernel image (ZO), has been moved so that its position
	390	* is against the end of the buffer used to hold the uncompressed kernel
	391	* image (VO) and the execution environment (.bss, .brk), which makes sure
	392	* there is room to do the in-place decompression. (See header.S for the
	393	* calculations.)
	394	*
	395	* \|-----compressed kernel image------\|
	396	* V V
	397	* 0 extract_offset +INIT_SIZE
	398	* \|-----------\|---------------\|-------------------------\|--------\|
	399	* \| \| \| \|
	400	* VO__text startup_32 of ZO VO__end ZO__end
	401	* ^ ^
	402	* \|-------uncompressed kernel image---------\|
	403	*
	404	*/
24388292	405	asmlinkage __visible void extract_kernel(void rmode, unsigned char *output)
1da177e4	406	{
8eabf42a	407	unsigned long virt_addr = LOAD_PHYSICAL_ADDR;
24388292	408	memptr heap = (memptr)boot_heap;
1869dbe8	409	unsigned long needed_size;
7734a0f3	410	size_t entry_offset;
f285f4a2	411
6655e0aa	412	/* Retain x86 boot parameters pointer passed from startup_32/64. */
d55d5bc5	413	boot_params_ptr = rmode;
1da177e4	414
6655e0aa	415	/* Clear flags intended for solely in-kernel use. */
d55d5bc5	416	boot_params_ptr->hdr.loadflags &= ~KASLR_FLAG;
78cac48c	417
cd0d9d92 AB	418	parse_mem_encrypt(&boot_params_ptr->hdr);
cd0d9d92 AB	419
d55d5bc5	420	sanitize_boot_params(boot_params_ptr);
5dcd14ec	421
d55d5bc5	422	if (boot_params_ptr->screen_info.orig_video_mode == 7) {
1da177e4 LT	423	vidmem = (char *) 0xb0000;
	424	vidport = 0x3b4;
	425	} else {
	426	vidmem = (char *) 0xb8000;
	427	vidport = 0x3d4;
	428	}
	429
d55d5bc5 AB	430	lines = boot_params_ptr->screen_info.orig_video_lines;
d55d5bc5 AB	431	cols = boot_params_ptr->screen_info.orig_video_cols;
1da177e4	432
eb4ea1ae KS	433	init_default_io_ops();
eb4ea1ae KS	434
4b05f815 KS	435	/*
	436	* Detect TDX guest environment.
	437	*
	438	* It has to be done before console_init() in order to use
	439	* paravirtualized port I/O operations if needed.
	440	*/
	441	early_tdx_detect();
	442
f30470c1 AK	443	early_sev_detect();
f30470c1 AK	444
8fee13a4	445	console_init();
5b51ae96 BP	446
	447	/*
	448	* Save RSDP address for later use. Have this after console_init()
	449	* so that early debugging output from the RSDP parsing code can be
	450	* collected.
	451	*/
d55d5bc5	452	boot_params_ptr->acpi_rsdp_addr = get_rsdp_addr();
5b51ae96	453
c0402881	454	debug_putstr("early console in extract_kernel\n");
8fee13a4	455
4c83d653	456	free_mem_ptr = heap; /* Heap */
7c539764	457	free_mem_end_ptr = heap + BOOT_HEAP_SIZE;
968de4f0	458
1869dbe8 SW	459	/*
	460	* The memory hole needed for the kernel is the larger of either
	461	* the entire decompressed kernel plus relocation table, or the
	462	* entire decompressed kernel plus .bss and .brk sections.
	463	*
	464	* On X86_64, the memory is mapped with PMD pages. Round the
	465	* size up so that the full extent of PMD pages mapped is
	466	* included in the check against the valid memory table
	467	* entries. This ensures the full mapped area is usable RAM
	468	* and doesn't include any reserved areas.
	469	*/
24388292	470	needed_size = max_t(unsigned long, output_len, kernel_total_size);
1869dbe8 SW	471	#ifdef CONFIG_X86_64
	472	needed_size = ALIGN(needed_size, MIN_KERNEL_ALIGN);
	473	#endif
	474
79063a7c KC	475	/* Report initial kernel position details. */
	476	debug_putaddr(input_data);
	477	debug_putaddr(input_len);
	478	debug_putaddr(output);
	479	debug_putaddr(output_len);
4d2d5424	480	debug_putaddr(kernel_total_size);
1869dbe8	481	debug_putaddr(needed_size);
79063a7c	482
3548e131 KS	483	#ifdef CONFIG_X86_64
	484	/* Report address of 32-bit trampoline */
	485	debug_putaddr(trampoline_32bit);
	486	#endif
	487
8391c73c BH	488	choose_random_location((unsigned long)input_data, input_len,
8391c73c BH	489	(unsigned long *)&output,
1869dbe8	490	needed_size,
8391c73c	491	&virt_addr);
8ab3820f KC	492
8ab3820f KC	493	/* Validate memory location choices. */
7ed42a28	494	if ((unsigned long)output & (MIN_KERNEL_ALIGN - 1))
8391c73c BH	495	error("Destination physical address inappropriately aligned");
	496	if (virt_addr & (MIN_KERNEL_ALIGN - 1))
	497	error("Destination virtual address inappropriately aligned");
778cb929	498	#ifdef CONFIG_X86_64
7ed42a28	499	if (heap > 0x3fffffffffffUL)
778cb929	500	error("Destination address too large");
24388292	501	if (virt_addr + needed_size > KERNEL_IMAGE_SIZE)
b892cb87	502	error("Destination virtual address is beyond the kernel mapping area");
778cb929	503	#else
147dd561	504	if (heap > ((-__PAGE_OFFSET-(128<<20)-1) & 0x7fffffff))
968de4f0	505	error("Destination address too large");
7ed42a28	506	#endif
968de4f0	507	#ifndef CONFIG_RELOCATABLE
8eabf42a	508	if (virt_addr != LOAD_PHYSICAL_ADDR)
8391c73c	509	error("Destination virtual address changed when not relocatable");
968de4f0	510	#endif
1da177e4	511
e605a425	512	debug_putstr("\nDecompressing Linux... ");
3fd1239a KS	513
	514	if (init_unaccepted_memory()) {
	515	debug_putstr("Accepting memory... ");
5adfeaec	516	accept_memory(__pa(output), needed_size);
3fd1239a KS	517	}
3fd1239a KS	518
83381519	519	entry_offset = decompress_kernel(output, virt_addr, error);
7734a0f3 AL	520
	521	debug_putstr("done.\nBooting the kernel (entry_offset: 0x");
	522	debug_puthex(entry_offset);
	523	debug_putstr(").\n");
597cfe48	524
b099155e JR	525	/* Disable exception handling before booting the kernel */
b099155e JR	526	cleanup_exception_handling();
597cfe48	527
ac456ca0 NJ	528	if (spurious_nmi_count) {
	529	error_putstr("Spurious early NMIs ignored: ");
	530	error_putdec(spurious_nmi_count);
	531	error_putstr("\n");
	532	}
	533
7734a0f3	534	return output + entry_offset;
1da177e4	535	}