kexec.c 53.4 KB
Newer Older
1 2 3 4 5 6 7 8
/*
 * kexec.c - kexec system call
 * Copyright (C) 2002-2004 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.
 */

9 10
#define pr_fmt(fmt)	"kexec: " fmt

11
#include <linux/capability.h>
12 13 14 15 16
#include <linux/mm.h>
#include <linux/file.h>
#include <linux/slab.h>
#include <linux/fs.h>
#include <linux/kexec.h>
17
#include <linux/mutex.h>
18 19 20 21 22
#include <linux/list.h>
#include <linux/highmem.h>
#include <linux/syscalls.h>
#include <linux/reboot.h>
#include <linux/ioport.h>
23
#include <linux/hardirq.h>
24 25
#include <linux/elf.h>
#include <linux/elfcore.h>
K
Ken'ichi Ohmichi 已提交
26 27
#include <linux/utsname.h>
#include <linux/numa.h>
H
Huang Ying 已提交
28 29
#include <linux/suspend.h>
#include <linux/device.h>
30 31 32 33
#include <linux/freezer.h>
#include <linux/pm.h>
#include <linux/cpu.h>
#include <linux/console.h>
34
#include <linux/vmalloc.h>
35
#include <linux/swap.h>
36
#include <linux/syscore_ops.h>
37
#include <linux/compiler.h>
38
#include <linux/hugetlb.h>
39

40 41 42
#include <asm/page.h>
#include <asm/uaccess.h>
#include <asm/io.h>
K
Ken'ichi Ohmichi 已提交
43
#include <asm/sections.h>
44

45
/* Per cpu memory for storing cpu states in case of system crash. */
46
note_buf_t __percpu *crash_notes;
47

K
Ken'ichi Ohmichi 已提交
48
/* vmcoreinfo stuff */
49
static unsigned char vmcoreinfo_data[VMCOREINFO_BYTES];
K
Ken'ichi Ohmichi 已提交
50
u32 vmcoreinfo_note[VMCOREINFO_NOTE_SIZE/4];
51 52
size_t vmcoreinfo_size;
size_t vmcoreinfo_max_size = sizeof(vmcoreinfo_data);
K
Ken'ichi Ohmichi 已提交
53

54 55 56
/* Flag to indicate we are going to kexec a new kernel */
bool kexec_in_progress = false;

57 58 59 60 61 62 63
/* Location of the reserved area for the crash kernel */
struct resource crashk_res = {
	.name  = "Crash kernel",
	.start = 0,
	.end   = 0,
	.flags = IORESOURCE_BUSY | IORESOURCE_MEM
};
64
struct resource crashk_low_res = {
65
	.name  = "Crash kernel",
66 67 68 69
	.start = 0,
	.end   = 0,
	.flags = IORESOURCE_BUSY | IORESOURCE_MEM
};
70

71 72
int kexec_should_crash(struct task_struct *p)
{
73
	if (in_interrupt() || !p->pid || is_global_init(p) || panic_on_oops)
74 75 76 77
		return 1;
	return 0;
}

78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93
/*
 * When kexec transitions to the new kernel there is a one-to-one
 * mapping between physical and virtual addresses.  On processors
 * where you can disable the MMU this is trivial, and easy.  For
 * others it is still a simple predictable page table to setup.
 *
 * In that environment kexec copies the new kernel to its final
 * resting place.  This means I can only support memory whose
 * physical address can fit in an unsigned long.  In particular
 * addresses where (pfn << PAGE_SHIFT) > ULONG_MAX cannot be handled.
 * If the assembly stub has more restrictive requirements
 * KEXEC_SOURCE_MEMORY_LIMIT and KEXEC_DEST_MEMORY_LIMIT can be
 * defined more restrictively in <asm/kexec.h>.
 *
 * The code for the transition from the current kernel to the
 * the new kernel is placed in the control_code_buffer, whose size
94
 * is given by KEXEC_CONTROL_PAGE_SIZE.  In the best case only a single
95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
 * page of memory is necessary, but some architectures require more.
 * Because this memory must be identity mapped in the transition from
 * virtual to physical addresses it must live in the range
 * 0 - TASK_SIZE, as only the user space mappings are arbitrarily
 * modifiable.
 *
 * The assembly stub in the control code buffer is passed a linked list
 * of descriptor pages detailing the source pages of the new kernel,
 * and the destination addresses of those source pages.  As this data
 * structure is not used in the context of the current OS, it must
 * be self-contained.
 *
 * The code has been made to work with highmem pages and will use a
 * destination page in its final resting place (if it happens
 * to allocate it).  The end product of this is that most of the
 * physical address space, and most of RAM can be used.
 *
 * Future directions include:
 *  - allocating a page table with the control code buffer identity
 *    mapped, to simplify machine_kexec and make kexec_on_panic more
 *    reliable.
 */

/*
 * KIMAGE_NO_DEST is an impossible destination address..., for
 * allocating pages whose destination address we do not care about.
 */
#define KIMAGE_NO_DEST (-1UL)

M
Maneesh Soni 已提交
124 125 126
static int kimage_is_destination_range(struct kimage *image,
				       unsigned long start, unsigned long end);
static struct page *kimage_alloc_page(struct kimage *image,
A
Al Viro 已提交
127
				       gfp_t gfp_mask,
M
Maneesh Soni 已提交
128
				       unsigned long dest);
129

130 131 132
static int copy_user_segment_list(struct kimage *image,
				  unsigned long nr_segments,
				  struct kexec_segment __user *segments)
133
{
134
	int ret;
135 136 137 138 139
	size_t segment_bytes;

	/* Read in the segments */
	image->nr_segments = nr_segments;
	segment_bytes = nr_segments * sizeof(*segments);
140 141 142 143 144 145 146 147 148 149 150
	ret = copy_from_user(image->segment, segments, segment_bytes);
	if (ret)
		ret = -EFAULT;

	return ret;
}

static int sanity_check_segment_list(struct kimage *image)
{
	int result, i;
	unsigned long nr_segments = image->nr_segments;
151 152 153 154 155 156 157 158

	/*
	 * Verify we have good destination addresses.  The caller is
	 * responsible for making certain we don't attempt to load
	 * the new image into invalid or reserved areas of RAM.  This
	 * just verifies it is an address we can use.
	 *
	 * Since the kernel does everything in page size chunks ensure
159
	 * the destination addresses are page aligned.  Too many
160 161 162 163 164 165 166 167
	 * special cases crop of when we don't do this.  The most
	 * insidious is getting overlapping destination addresses
	 * simply because addresses are changed to page size
	 * granularity.
	 */
	result = -EADDRNOTAVAIL;
	for (i = 0; i < nr_segments; i++) {
		unsigned long mstart, mend;
M
Maneesh Soni 已提交
168

169 170 171
		mstart = image->segment[i].mem;
		mend   = mstart + image->segment[i].memsz;
		if ((mstart & ~PAGE_MASK) || (mend & ~PAGE_MASK))
172
			return result;
173
		if (mend >= KEXEC_DESTINATION_MEMORY_LIMIT)
174
			return result;
175 176 177 178 179 180 181 182
	}

	/* Verify our destination addresses do not overlap.
	 * If we alloed overlapping destination addresses
	 * through very weird things can happen with no
	 * easy explanation as one segment stops on another.
	 */
	result = -EINVAL;
M
Maneesh Soni 已提交
183
	for (i = 0; i < nr_segments; i++) {
184 185
		unsigned long mstart, mend;
		unsigned long j;
M
Maneesh Soni 已提交
186

187 188
		mstart = image->segment[i].mem;
		mend   = mstart + image->segment[i].memsz;
M
Maneesh Soni 已提交
189
		for (j = 0; j < i; j++) {
190 191 192 193 194
			unsigned long pstart, pend;
			pstart = image->segment[j].mem;
			pend   = pstart + image->segment[j].memsz;
			/* Do the segments overlap ? */
			if ((mend > pstart) && (mstart < pend))
195
				return result;
196 197 198 199 200 201 202 203 204
		}
	}

	/* Ensure our buffer sizes are strictly less than
	 * our memory sizes.  This should always be the case,
	 * and it is easier to check up front than to be surprised
	 * later on.
	 */
	result = -EINVAL;
M
Maneesh Soni 已提交
205
	for (i = 0; i < nr_segments; i++) {
206
		if (image->segment[i].bufsz > image->segment[i].memsz)
207
			return result;
208 209
	}

210 211 212 213 214 215 216 217 218
	/*
	 * Verify we have good destination addresses.  Normally
	 * the caller is responsible for making certain we don't
	 * attempt to load the new image into invalid or reserved
	 * areas of RAM.  But crash kernels are preloaded into a
	 * reserved area of ram.  We must ensure the addresses
	 * are in the reserved area otherwise preloading the
	 * kernel could corrupt things.
	 */
M
Maneesh Soni 已提交
219

220 221 222 223 224 225 226 227 228 229 230 231 232
	if (image->type == KEXEC_TYPE_CRASH) {
		result = -EADDRNOTAVAIL;
		for (i = 0; i < nr_segments; i++) {
			unsigned long mstart, mend;

			mstart = image->segment[i].mem;
			mend = mstart + image->segment[i].memsz - 1;
			/* Ensure we are within the crash kernel limits */
			if ((mstart < crashk_res.start) ||
			    (mend > crashk_res.end))
				return result;
		}
	}
233

234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261
	return 0;
}

static struct kimage *do_kimage_alloc_init(void)
{
	struct kimage *image;

	/* Allocate a controlling structure */
	image = kzalloc(sizeof(*image), GFP_KERNEL);
	if (!image)
		return NULL;

	image->head = 0;
	image->entry = &image->head;
	image->last_entry = &image->head;
	image->control_page = ~0; /* By default this does not apply */
	image->type = KEXEC_TYPE_DEFAULT;

	/* Initialize the list of control pages */
	INIT_LIST_HEAD(&image->control_pages);

	/* Initialize the list of destination pages */
	INIT_LIST_HEAD(&image->dest_pages);

	/* Initialize the list of unusable pages */
	INIT_LIST_HEAD(&image->unusable_pages);

	return image;
262 263
}

264 265
static void kimage_free_page_list(struct list_head *list);

266 267 268 269
static int kimage_alloc_init(struct kimage **rimage, unsigned long entry,
			     unsigned long nr_segments,
			     struct kexec_segment __user *segments,
			     unsigned long flags)
270
{
271
	int ret;
272
	struct kimage *image;
273 274 275 276 277 278 279
	bool kexec_on_panic = flags & KEXEC_ON_CRASH;

	if (kexec_on_panic) {
		/* Verify we have a valid entry point */
		if ((entry < crashk_res.start) || (entry > crashk_res.end))
			return -EADDRNOTAVAIL;
	}
280 281

	/* Allocate and initialize a controlling structure */
282 283 284 285 286 287
	image = do_kimage_alloc_init();
	if (!image)
		return -ENOMEM;

	image->start = entry;

288 289
	ret = copy_user_segment_list(image, nr_segments, segments);
	if (ret)
290 291
		goto out_free_image;

292 293
	ret = sanity_check_segment_list(image);
	if (ret)
294
		goto out_free_image;
M
Maneesh Soni 已提交
295

296 297 298 299 300 301
	 /* Enable the special crash kernel control page allocation policy. */
	if (kexec_on_panic) {
		image->control_page = crashk_res.start;
		image->type = KEXEC_TYPE_CRASH;
	}

302 303 304 305 306
	/*
	 * Find a location for the control code buffer, and add it
	 * the vector of segments so that it's pages will also be
	 * counted as destination pages.
	 */
307
	ret = -ENOMEM;
308
	image->control_code_page = kimage_alloc_control_pages(image,
309
					   get_order(KEXEC_CONTROL_PAGE_SIZE));
310
	if (!image->control_code_page) {
311
		pr_err("Could not allocate control_code_buffer\n");
312
		goto out_free_image;
313 314
	}

315 316 317 318 319 320
	if (!kexec_on_panic) {
		image->swap_page = kimage_alloc_control_pages(image, 0);
		if (!image->swap_page) {
			pr_err("Could not allocate swap buffer\n");
			goto out_free_control_pages;
		}
H
Huang Ying 已提交
321 322
	}

323 324
	*rimage = image;
	return 0;
325
out_free_control_pages:
326
	kimage_free_page_list(&image->control_pages);
327
out_free_image:
328
	kfree(image);
329
	return ret;
330 331
}

332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546
static int copy_file_from_fd(int fd, void **buf, unsigned long *buf_len)
{
	struct fd f = fdget(fd);
	int ret;
	struct kstat stat;
	loff_t pos;
	ssize_t bytes = 0;

	if (!f.file)
		return -EBADF;

	ret = vfs_getattr(&f.file->f_path, &stat);
	if (ret)
		goto out;

	if (stat.size > INT_MAX) {
		ret = -EFBIG;
		goto out;
	}

	/* Don't hand 0 to vmalloc, it whines. */
	if (stat.size == 0) {
		ret = -EINVAL;
		goto out;
	}

	*buf = vmalloc(stat.size);
	if (!*buf) {
		ret = -ENOMEM;
		goto out;
	}

	pos = 0;
	while (pos < stat.size) {
		bytes = kernel_read(f.file, pos, (char *)(*buf) + pos,
				    stat.size - pos);
		if (bytes < 0) {
			vfree(*buf);
			ret = bytes;
			goto out;
		}

		if (bytes == 0)
			break;
		pos += bytes;
	}

	if (pos != stat.size) {
		ret = -EBADF;
		vfree(*buf);
		goto out;
	}

	*buf_len = pos;
out:
	fdput(f);
	return ret;
}

/* Architectures can provide this probe function */
int __weak arch_kexec_kernel_image_probe(struct kimage *image, void *buf,
					 unsigned long buf_len)
{
	return -ENOEXEC;
}

void * __weak arch_kexec_kernel_image_load(struct kimage *image)
{
	return ERR_PTR(-ENOEXEC);
}

void __weak arch_kimage_file_post_load_cleanup(struct kimage *image)
{
}

/*
 * Free up memory used by kernel, initrd, and comand line. This is temporary
 * memory allocation which is not needed any more after these buffers have
 * been loaded into separate segments and have been copied elsewhere.
 */
static void kimage_file_post_load_cleanup(struct kimage *image)
{
	vfree(image->kernel_buf);
	image->kernel_buf = NULL;

	vfree(image->initrd_buf);
	image->initrd_buf = NULL;

	kfree(image->cmdline_buf);
	image->cmdline_buf = NULL;

	/* See if architecture has anything to cleanup post load */
	arch_kimage_file_post_load_cleanup(image);
}

/*
 * In file mode list of segments is prepared by kernel. Copy relevant
 * data from user space, do error checking, prepare segment list
 */
static int
kimage_file_prepare_segments(struct kimage *image, int kernel_fd, int initrd_fd,
			     const char __user *cmdline_ptr,
			     unsigned long cmdline_len, unsigned flags)
{
	int ret = 0;
	void *ldata;

	ret = copy_file_from_fd(kernel_fd, &image->kernel_buf,
				&image->kernel_buf_len);
	if (ret)
		return ret;

	/* Call arch image probe handlers */
	ret = arch_kexec_kernel_image_probe(image, image->kernel_buf,
					    image->kernel_buf_len);

	if (ret)
		goto out;

	/* It is possible that there no initramfs is being loaded */
	if (!(flags & KEXEC_FILE_NO_INITRAMFS)) {
		ret = copy_file_from_fd(initrd_fd, &image->initrd_buf,
					&image->initrd_buf_len);
		if (ret)
			goto out;
	}

	if (cmdline_len) {
		image->cmdline_buf = kzalloc(cmdline_len, GFP_KERNEL);
		if (!image->cmdline_buf) {
			ret = -ENOMEM;
			goto out;
		}

		ret = copy_from_user(image->cmdline_buf, cmdline_ptr,
				     cmdline_len);
		if (ret) {
			ret = -EFAULT;
			goto out;
		}

		image->cmdline_buf_len = cmdline_len;

		/* command line should be a string with last byte null */
		if (image->cmdline_buf[cmdline_len - 1] != '\0') {
			ret = -EINVAL;
			goto out;
		}
	}

	/* Call arch image load handlers */
	ldata = arch_kexec_kernel_image_load(image);

	if (IS_ERR(ldata)) {
		ret = PTR_ERR(ldata);
		goto out;
	}

	image->image_loader_data = ldata;
out:
	/* In case of error, free up all allocated memory in this function */
	if (ret)
		kimage_file_post_load_cleanup(image);
	return ret;
}

static int
kimage_file_alloc_init(struct kimage **rimage, int kernel_fd,
		       int initrd_fd, const char __user *cmdline_ptr,
		       unsigned long cmdline_len, unsigned long flags)
{
	int ret;
	struct kimage *image;

	image = do_kimage_alloc_init();
	if (!image)
		return -ENOMEM;

	image->file_mode = 1;

	ret = kimage_file_prepare_segments(image, kernel_fd, initrd_fd,
					   cmdline_ptr, cmdline_len, flags);
	if (ret)
		goto out_free_image;

	ret = sanity_check_segment_list(image);
	if (ret)
		goto out_free_post_load_bufs;

	ret = -ENOMEM;
	image->control_code_page = kimage_alloc_control_pages(image,
					   get_order(KEXEC_CONTROL_PAGE_SIZE));
	if (!image->control_code_page) {
		pr_err("Could not allocate control_code_buffer\n");
		goto out_free_post_load_bufs;
	}

	image->swap_page = kimage_alloc_control_pages(image, 0);
	if (!image->swap_page) {
		pr_err(KERN_ERR "Could not allocate swap buffer\n");
		goto out_free_control_pages;
	}

	*rimage = image;
	return 0;
out_free_control_pages:
	kimage_free_page_list(&image->control_pages);
out_free_post_load_bufs:
	kimage_file_post_load_cleanup(image);
	kfree(image->image_loader_data);
out_free_image:
	kfree(image);
	return ret;
}

M
Maneesh Soni 已提交
547 548 549
static int kimage_is_destination_range(struct kimage *image,
					unsigned long start,
					unsigned long end)
550 551 552 553 554
{
	unsigned long i;

	for (i = 0; i < image->nr_segments; i++) {
		unsigned long mstart, mend;
M
Maneesh Soni 已提交
555

556
		mstart = image->segment[i].mem;
M
Maneesh Soni 已提交
557 558
		mend = mstart + image->segment[i].memsz;
		if ((end > mstart) && (start < mend))
559 560
			return 1;
	}
M
Maneesh Soni 已提交
561

562 563 564
	return 0;
}

A
Al Viro 已提交
565
static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order)
566 567
{
	struct page *pages;
M
Maneesh Soni 已提交
568

569 570 571 572
	pages = alloc_pages(gfp_mask, order);
	if (pages) {
		unsigned int count, i;
		pages->mapping = NULL;
H
Hugh Dickins 已提交
573
		set_page_private(pages, order);
574
		count = 1 << order;
M
Maneesh Soni 已提交
575
		for (i = 0; i < count; i++)
576 577
			SetPageReserved(pages + i);
	}
M
Maneesh Soni 已提交
578

579 580 581 582 583 584
	return pages;
}

static void kimage_free_pages(struct page *page)
{
	unsigned int order, count, i;
M
Maneesh Soni 已提交
585

H
Hugh Dickins 已提交
586
	order = page_private(page);
587
	count = 1 << order;
M
Maneesh Soni 已提交
588
	for (i = 0; i < count; i++)
589 590 591 592 593 594 595
		ClearPageReserved(page + i);
	__free_pages(page, order);
}

static void kimage_free_page_list(struct list_head *list)
{
	struct list_head *pos, *next;
M
Maneesh Soni 已提交
596

597 598 599 600 601 602 603 604 605
	list_for_each_safe(pos, next, list) {
		struct page *page;

		page = list_entry(pos, struct page, lru);
		list_del(&page->lru);
		kimage_free_pages(page);
	}
}

M
Maneesh Soni 已提交
606 607
static struct page *kimage_alloc_normal_control_pages(struct kimage *image,
							unsigned int order)
608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633
{
	/* Control pages are special, they are the intermediaries
	 * that are needed while we copy the rest of the pages
	 * to their final resting place.  As such they must
	 * not conflict with either the destination addresses
	 * or memory the kernel is already using.
	 *
	 * The only case where we really need more than one of
	 * these are for architectures where we cannot disable
	 * the MMU and must instead generate an identity mapped
	 * page table for all of the memory.
	 *
	 * At worst this runs in O(N) of the image size.
	 */
	struct list_head extra_pages;
	struct page *pages;
	unsigned int count;

	count = 1 << order;
	INIT_LIST_HEAD(&extra_pages);

	/* Loop while I can allocate a page and the page allocated
	 * is a destination page.
	 */
	do {
		unsigned long pfn, epfn, addr, eaddr;
M
Maneesh Soni 已提交
634

635 636 637 638 639 640 641 642
		pages = kimage_alloc_pages(GFP_KERNEL, order);
		if (!pages)
			break;
		pfn   = page_to_pfn(pages);
		epfn  = pfn + count;
		addr  = pfn << PAGE_SHIFT;
		eaddr = epfn << PAGE_SHIFT;
		if ((epfn >= (KEXEC_CONTROL_MEMORY_LIMIT >> PAGE_SHIFT)) ||
M
Maneesh Soni 已提交
643
			      kimage_is_destination_range(image, addr, eaddr)) {
644 645 646
			list_add(&pages->lru, &extra_pages);
			pages = NULL;
		}
M
Maneesh Soni 已提交
647 648
	} while (!pages);

649 650 651 652 653 654 655 656 657 658 659 660 661 662
	if (pages) {
		/* Remember the allocated page... */
		list_add(&pages->lru, &image->control_pages);

		/* Because the page is already in it's destination
		 * location we will never allocate another page at
		 * that address.  Therefore kimage_alloc_pages
		 * will not return it (again) and we don't need
		 * to give it an entry in image->segment[].
		 */
	}
	/* Deal with the destination pages I have inadvertently allocated.
	 *
	 * Ideally I would convert multi-page allocations into single
L
Lucas De Marchi 已提交
663
	 * page allocations, and add everything to image->dest_pages.
664 665 666 667 668
	 *
	 * For now it is simpler to just free the pages.
	 */
	kimage_free_page_list(&extra_pages);

M
Maneesh Soni 已提交
669
	return pages;
670 671
}

M
Maneesh Soni 已提交
672 673
static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
						      unsigned int order)
674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697
{
	/* Control pages are special, they are the intermediaries
	 * that are needed while we copy the rest of the pages
	 * to their final resting place.  As such they must
	 * not conflict with either the destination addresses
	 * or memory the kernel is already using.
	 *
	 * Control pages are also the only pags we must allocate
	 * when loading a crash kernel.  All of the other pages
	 * are specified by the segments and we just memcpy
	 * into them directly.
	 *
	 * The only case where we really need more than one of
	 * these are for architectures where we cannot disable
	 * the MMU and must instead generate an identity mapped
	 * page table for all of the memory.
	 *
	 * Given the low demand this implements a very simple
	 * allocator that finds the first hole of the appropriate
	 * size in the reserved memory region, and allocates all
	 * of the memory up to and including the hole.
	 */
	unsigned long hole_start, hole_end, size;
	struct page *pages;
M
Maneesh Soni 已提交
698

699 700 701 702
	pages = NULL;
	size = (1 << order) << PAGE_SHIFT;
	hole_start = (image->control_page + (size - 1)) & ~(size - 1);
	hole_end   = hole_start + size - 1;
M
Maneesh Soni 已提交
703
	while (hole_end <= crashk_res.end) {
704
		unsigned long i;
M
Maneesh Soni 已提交
705

706
		if (hole_end > KEXEC_CRASH_CONTROL_MEMORY_LIMIT)
707 708
			break;
		/* See if I overlap any of the segments */
M
Maneesh Soni 已提交
709
		for (i = 0; i < image->nr_segments; i++) {
710
			unsigned long mstart, mend;
M
Maneesh Soni 已提交
711

712 713 714 715 716 717 718 719 720 721 722 723 724 725 726
			mstart = image->segment[i].mem;
			mend   = mstart + image->segment[i].memsz - 1;
			if ((hole_end >= mstart) && (hole_start <= mend)) {
				/* Advance the hole to the end of the segment */
				hole_start = (mend + (size - 1)) & ~(size - 1);
				hole_end   = hole_start + size - 1;
				break;
			}
		}
		/* If I don't overlap any segments I have found my hole! */
		if (i == image->nr_segments) {
			pages = pfn_to_page(hole_start >> PAGE_SHIFT);
			break;
		}
	}
M
Maneesh Soni 已提交
727
	if (pages)
728
		image->control_page = hole_end;
M
Maneesh Soni 已提交
729

730 731 732 733
	return pages;
}


M
Maneesh Soni 已提交
734 735
struct page *kimage_alloc_control_pages(struct kimage *image,
					 unsigned int order)
736 737
{
	struct page *pages = NULL;
M
Maneesh Soni 已提交
738 739

	switch (image->type) {
740 741 742 743 744 745 746
	case KEXEC_TYPE_DEFAULT:
		pages = kimage_alloc_normal_control_pages(image, order);
		break;
	case KEXEC_TYPE_CRASH:
		pages = kimage_alloc_crash_control_pages(image, order);
		break;
	}
M
Maneesh Soni 已提交
747

748 749 750 751 752
	return pages;
}

static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
{
M
Maneesh Soni 已提交
753
	if (*image->entry != 0)
754
		image->entry++;
M
Maneesh Soni 已提交
755

756 757 758
	if (image->entry == image->last_entry) {
		kimage_entry_t *ind_page;
		struct page *page;
M
Maneesh Soni 已提交
759

760
		page = kimage_alloc_page(image, GFP_KERNEL, KIMAGE_NO_DEST);
M
Maneesh Soni 已提交
761
		if (!page)
762
			return -ENOMEM;
M
Maneesh Soni 已提交
763

764 765 766
		ind_page = page_address(page);
		*image->entry = virt_to_phys(ind_page) | IND_INDIRECTION;
		image->entry = ind_page;
M
Maneesh Soni 已提交
767 768
		image->last_entry = ind_page +
				      ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1);
769 770 771 772
	}
	*image->entry = entry;
	image->entry++;
	*image->entry = 0;
M
Maneesh Soni 已提交
773

774 775 776
	return 0;
}

M
Maneesh Soni 已提交
777 778
static int kimage_set_destination(struct kimage *image,
				   unsigned long destination)
779 780 781 782 783
{
	int result;

	destination &= PAGE_MASK;
	result = kimage_add_entry(image, destination | IND_DESTINATION);
M
Maneesh Soni 已提交
784
	if (result == 0)
785
		image->destination = destination;
M
Maneesh Soni 已提交
786

787 788 789 790 791 792 793 794 795 796
	return result;
}


static int kimage_add_page(struct kimage *image, unsigned long page)
{
	int result;

	page &= PAGE_MASK;
	result = kimage_add_entry(image, page | IND_SOURCE);
M
Maneesh Soni 已提交
797
	if (result == 0)
798
		image->destination += PAGE_SIZE;
M
Maneesh Soni 已提交
799

800 801 802 803 804 805 806 807 808
	return result;
}


static void kimage_free_extra_pages(struct kimage *image)
{
	/* Walk through and free any extra destination pages I may have */
	kimage_free_page_list(&image->dest_pages);

L
Lucas De Marchi 已提交
809
	/* Walk through and free any unusable pages I have cached */
810
	kimage_free_page_list(&image->unusable_pages);
811 812

}
813
static void kimage_terminate(struct kimage *image)
814
{
M
Maneesh Soni 已提交
815
	if (*image->entry != 0)
816
		image->entry++;
M
Maneesh Soni 已提交
817

818 819 820 821 822
	*image->entry = IND_DONE;
}

#define for_each_kimage_entry(image, ptr, entry) \
	for (ptr = &image->head; (entry = *ptr) && !(entry & IND_DONE); \
823 824
		ptr = (entry & IND_INDIRECTION) ? \
			phys_to_virt((entry & PAGE_MASK)) : ptr + 1)
825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840

static void kimage_free_entry(kimage_entry_t entry)
{
	struct page *page;

	page = pfn_to_page(entry >> PAGE_SHIFT);
	kimage_free_pages(page);
}

static void kimage_free(struct kimage *image)
{
	kimage_entry_t *ptr, entry;
	kimage_entry_t ind = 0;

	if (!image)
		return;
M
Maneesh Soni 已提交
841

842 843 844 845
	kimage_free_extra_pages(image);
	for_each_kimage_entry(image, ptr, entry) {
		if (entry & IND_INDIRECTION) {
			/* Free the previous indirection page */
M
Maneesh Soni 已提交
846
			if (ind & IND_INDIRECTION)
847 848 849 850 851
				kimage_free_entry(ind);
			/* Save this indirection page until we are
			 * done with it.
			 */
			ind = entry;
852
		} else if (entry & IND_SOURCE)
853 854 855
			kimage_free_entry(entry);
	}
	/* Free the final indirection page */
M
Maneesh Soni 已提交
856
	if (ind & IND_INDIRECTION)
857 858 859 860 861 862 863
		kimage_free_entry(ind);

	/* Handle any machine specific cleanup */
	machine_kexec_cleanup(image);

	/* Free the kexec control pages... */
	kimage_free_page_list(&image->control_pages);
864 865 866 867 868 869 870 871 872 873

	kfree(image->image_loader_data);

	/*
	 * Free up any temporary buffers allocated. This might hit if
	 * error occurred much later after buffer allocation.
	 */
	if (image->file_mode)
		kimage_file_post_load_cleanup(image);

874 875 876
	kfree(image);
}

M
Maneesh Soni 已提交
877 878
static kimage_entry_t *kimage_dst_used(struct kimage *image,
					unsigned long page)
879 880 881 882 883
{
	kimage_entry_t *ptr, entry;
	unsigned long destination = 0;

	for_each_kimage_entry(image, ptr, entry) {
M
Maneesh Soni 已提交
884
		if (entry & IND_DESTINATION)
885 886
			destination = entry & PAGE_MASK;
		else if (entry & IND_SOURCE) {
M
Maneesh Soni 已提交
887
			if (page == destination)
888 889 890 891
				return ptr;
			destination += PAGE_SIZE;
		}
	}
M
Maneesh Soni 已提交
892

893
	return NULL;
894 895
}

M
Maneesh Soni 已提交
896
static struct page *kimage_alloc_page(struct kimage *image,
A
Al Viro 已提交
897
					gfp_t gfp_mask,
M
Maneesh Soni 已提交
898
					unsigned long destination)
899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937
{
	/*
	 * Here we implement safeguards to ensure that a source page
	 * is not copied to its destination page before the data on
	 * the destination page is no longer useful.
	 *
	 * To do this we maintain the invariant that a source page is
	 * either its own destination page, or it is not a
	 * destination page at all.
	 *
	 * That is slightly stronger than required, but the proof
	 * that no problems will not occur is trivial, and the
	 * implementation is simply to verify.
	 *
	 * When allocating all pages normally this algorithm will run
	 * in O(N) time, but in the worst case it will run in O(N^2)
	 * time.   If the runtime is a problem the data structures can
	 * be fixed.
	 */
	struct page *page;
	unsigned long addr;

	/*
	 * Walk through the list of destination pages, and see if I
	 * have a match.
	 */
	list_for_each_entry(page, &image->dest_pages, lru) {
		addr = page_to_pfn(page) << PAGE_SHIFT;
		if (addr == destination) {
			list_del(&page->lru);
			return page;
		}
	}
	page = NULL;
	while (1) {
		kimage_entry_t *old;

		/* Allocate a page, if we run out of memory give up */
		page = kimage_alloc_pages(gfp_mask, 0);
M
Maneesh Soni 已提交
938
		if (!page)
939
			return NULL;
940
		/* If the page cannot be used file it away */
M
Maneesh Soni 已提交
941 942
		if (page_to_pfn(page) >
				(KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) {
943
			list_add(&page->lru, &image->unusable_pages);
944 945 946 947 948 949 950 951 952
			continue;
		}
		addr = page_to_pfn(page) << PAGE_SHIFT;

		/* If it is the destination page we want use it */
		if (addr == destination)
			break;

		/* If the page is not a destination page use it */
M
Maneesh Soni 已提交
953 954
		if (!kimage_is_destination_range(image, addr,
						  addr + PAGE_SIZE))
955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973
			break;

		/*
		 * I know that the page is someones destination page.
		 * See if there is already a source page for this
		 * destination page.  And if so swap the source pages.
		 */
		old = kimage_dst_used(image, addr);
		if (old) {
			/* If so move it */
			unsigned long old_addr;
			struct page *old_page;

			old_addr = *old & PAGE_MASK;
			old_page = pfn_to_page(old_addr >> PAGE_SHIFT);
			copy_highpage(page, old_page);
			*old = addr | (*old & ~PAGE_MASK);

			/* The old page I have found cannot be a
974 975
			 * destination page, so return it if it's
			 * gfp_flags honor the ones passed in.
976
			 */
977 978 979 980 981
			if (!(gfp_mask & __GFP_HIGHMEM) &&
			    PageHighMem(old_page)) {
				kimage_free_pages(old_page);
				continue;
			}
982 983 984
			addr = old_addr;
			page = old_page;
			break;
985
		} else {
986 987 988 989 990 991
			/* Place the page on the destination list I
			 * will use it later.
			 */
			list_add(&page->lru, &image->dest_pages);
		}
	}
M
Maneesh Soni 已提交
992

993 994 995 996
	return page;
}

static int kimage_load_normal_segment(struct kimage *image,
M
Maneesh Soni 已提交
997
					 struct kexec_segment *segment)
998 999
{
	unsigned long maddr;
1000
	size_t ubytes, mbytes;
1001
	int result;
1002 1003
	unsigned char __user *buf = NULL;
	unsigned char *kbuf = NULL;
1004 1005

	result = 0;
1006 1007 1008 1009
	if (image->file_mode)
		kbuf = segment->kbuf;
	else
		buf = segment->buf;
1010 1011 1012 1013 1014
	ubytes = segment->bufsz;
	mbytes = segment->memsz;
	maddr = segment->mem;

	result = kimage_set_destination(image, maddr);
M
Maneesh Soni 已提交
1015
	if (result < 0)
1016
		goto out;
M
Maneesh Soni 已提交
1017 1018

	while (mbytes) {
1019 1020 1021
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
1022

1023
		page = kimage_alloc_page(image, GFP_HIGHUSER, maddr);
1024
		if (!page) {
1025 1026 1027
			result  = -ENOMEM;
			goto out;
		}
M
Maneesh Soni 已提交
1028 1029 1030
		result = kimage_add_page(image, page_to_pfn(page)
								<< PAGE_SHIFT);
		if (result < 0)
1031
			goto out;
M
Maneesh Soni 已提交
1032

1033 1034
		ptr = kmap(page);
		/* Start with a clear page */
1035
		clear_page(ptr);
1036
		ptr += maddr & ~PAGE_MASK;
1037 1038 1039
		mchunk = min_t(size_t, mbytes,
				PAGE_SIZE - (maddr & ~PAGE_MASK));
		uchunk = min(ubytes, mchunk);
M
Maneesh Soni 已提交
1040

1041 1042 1043 1044 1045
		/* For file based kexec, source pages are in kernel memory */
		if (image->file_mode)
			memcpy(ptr, kbuf, uchunk);
		else
			result = copy_from_user(ptr, buf, uchunk);
1046 1047
		kunmap(page);
		if (result) {
1048
			result = -EFAULT;
1049 1050 1051 1052
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
1053 1054 1055 1056
		if (image->file_mode)
			kbuf += mchunk;
		else
			buf += mchunk;
1057 1058
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
1059
out:
1060 1061 1062 1063
	return result;
}

static int kimage_load_crash_segment(struct kimage *image,
M
Maneesh Soni 已提交
1064
					struct kexec_segment *segment)
1065 1066 1067 1068 1069 1070
{
	/* For crash dumps kernels we simply copy the data from
	 * user space to it's destination.
	 * We do things a page at a time for the sake of kmap.
	 */
	unsigned long maddr;
1071
	size_t ubytes, mbytes;
1072
	int result;
1073
	unsigned char __user *buf;
1074 1075 1076 1077 1078 1079

	result = 0;
	buf = segment->buf;
	ubytes = segment->bufsz;
	mbytes = segment->memsz;
	maddr = segment->mem;
M
Maneesh Soni 已提交
1080
	while (mbytes) {
1081 1082 1083
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
1084

1085
		page = pfn_to_page(maddr >> PAGE_SHIFT);
1086
		if (!page) {
1087 1088 1089 1090 1091
			result  = -ENOMEM;
			goto out;
		}
		ptr = kmap(page);
		ptr += maddr & ~PAGE_MASK;
1092 1093 1094 1095
		mchunk = min_t(size_t, mbytes,
				PAGE_SIZE - (maddr & ~PAGE_MASK));
		uchunk = min(ubytes, mchunk);
		if (mchunk > uchunk) {
1096 1097 1098 1099
			/* Zero the trailing part of the page */
			memset(ptr + uchunk, 0, mchunk - uchunk);
		}
		result = copy_from_user(ptr, buf, uchunk);
Z
Zou Nan hai 已提交
1100
		kexec_flush_icache_page(page);
1101 1102
		kunmap(page);
		if (result) {
1103
			result = -EFAULT;
1104 1105 1106 1107 1108 1109 1110
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
		buf    += mchunk;
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
1111
out:
1112 1113 1114 1115
	return result;
}

static int kimage_load_segment(struct kimage *image,
M
Maneesh Soni 已提交
1116
				struct kexec_segment *segment)
1117 1118
{
	int result = -ENOMEM;
M
Maneesh Soni 已提交
1119 1120

	switch (image->type) {
1121 1122 1123 1124 1125 1126 1127
	case KEXEC_TYPE_DEFAULT:
		result = kimage_load_normal_segment(image, segment);
		break;
	case KEXEC_TYPE_CRASH:
		result = kimage_load_crash_segment(image, segment);
		break;
	}
M
Maneesh Soni 已提交
1128

1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145
	return result;
}

/*
 * Exec Kernel system call: for obvious reasons only root may call it.
 *
 * This call breaks up into three pieces.
 * - A generic part which loads the new kernel from the current
 *   address space, and very carefully places the data in the
 *   allocated pages.
 *
 * - A generic part that interacts with the kernel and tells all of
 *   the devices to shut down.  Preventing on-going dmas, and placing
 *   the devices in a consistent state so a later kernel can
 *   reinitialize them.
 *
 * - A machine specific part that includes the syscall number
G
Geert Uytterhoeven 已提交
1146
 *   and then copies the image to it's final destination.  And
1147 1148 1149 1150 1151
 *   jumps into the image at entry.
 *
 * kexec does not sync, or unmount filesystems so if you need
 * that to happen you need to do that yourself.
 */
1152 1153
struct kimage *kexec_image;
struct kimage *kexec_crash_image;
1154
int kexec_load_disabled;
1155 1156

static DEFINE_MUTEX(kexec_mutex);
1157

1158 1159
SYSCALL_DEFINE4(kexec_load, unsigned long, entry, unsigned long, nr_segments,
		struct kexec_segment __user *, segments, unsigned long, flags)
1160 1161 1162 1163 1164
{
	struct kimage **dest_image, *image;
	int result;

	/* We only trust the superuser with rebooting the system. */
1165
	if (!capable(CAP_SYS_BOOT) || kexec_load_disabled)
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
		return -EPERM;

	/*
	 * Verify we have a legal set of flags
	 * This leaves us room for future extensions.
	 */
	if ((flags & KEXEC_FLAGS) != (flags & ~KEXEC_ARCH_MASK))
		return -EINVAL;

	/* Verify we are on the appropriate architecture */
	if (((flags & KEXEC_ARCH_MASK) != KEXEC_ARCH) &&
		((flags & KEXEC_ARCH_MASK) != KEXEC_ARCH_DEFAULT))
		return -EINVAL;

	/* Put an artificial cap on the number
	 * of segments passed to kexec_load.
	 */
	if (nr_segments > KEXEC_SEGMENT_MAX)
		return -EINVAL;

	image = NULL;
	result = 0;

	/* Because we write directly to the reserved memory
	 * region when loading crash kernels we need a mutex here to
	 * prevent multiple crash  kernels from attempting to load
	 * simultaneously, and to prevent a crash kernel from loading
	 * over the top of a in use crash kernel.
	 *
	 * KISS: always take the mutex.
	 */
1197
	if (!mutex_trylock(&kexec_mutex))
1198
		return -EBUSY;
M
Maneesh Soni 已提交
1199

1200
	dest_image = &kexec_image;
M
Maneesh Soni 已提交
1201
	if (flags & KEXEC_ON_CRASH)
1202 1203 1204
		dest_image = &kexec_crash_image;
	if (nr_segments > 0) {
		unsigned long i;
M
Maneesh Soni 已提交
1205

1206
		/* Loading another kernel to reboot into */
M
Maneesh Soni 已提交
1207
		if ((flags & KEXEC_ON_CRASH) == 0)
1208 1209
			result = kimage_alloc_init(&image, entry, nr_segments,
						   segments, flags);
1210 1211 1212 1213 1214 1215
		/* Loading another kernel to switch to if this one crashes */
		else if (flags & KEXEC_ON_CRASH) {
			/* Free any current crash dump kernel before
			 * we corrupt it.
			 */
			kimage_free(xchg(&kexec_crash_image, NULL));
1216 1217
			result = kimage_alloc_init(&image, entry, nr_segments,
						   segments, flags);
1218
			crash_map_reserved_pages();
1219
		}
M
Maneesh Soni 已提交
1220
		if (result)
1221
			goto out;
M
Maneesh Soni 已提交
1222

H
Huang Ying 已提交
1223 1224
		if (flags & KEXEC_PRESERVE_CONTEXT)
			image->preserve_context = 1;
1225
		result = machine_kexec_prepare(image);
M
Maneesh Soni 已提交
1226
		if (result)
1227
			goto out;
M
Maneesh Soni 已提交
1228 1229

		for (i = 0; i < nr_segments; i++) {
1230
			result = kimage_load_segment(image, &image->segment[i]);
M
Maneesh Soni 已提交
1231
			if (result)
1232 1233
				goto out;
		}
1234
		kimage_terminate(image);
1235 1236
		if (flags & KEXEC_ON_CRASH)
			crash_unmap_reserved_pages();
1237 1238 1239 1240
	}
	/* Install the new kernel, and  Uninstall the old */
	image = xchg(dest_image, image);

M
Maneesh Soni 已提交
1241
out:
1242
	mutex_unlock(&kexec_mutex);
1243
	kimage_free(image);
M
Maneesh Soni 已提交
1244

1245 1246 1247
	return result;
}

1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259
/*
 * Add and remove page tables for crashkernel memory
 *
 * Provide an empty default implementation here -- architecture
 * code may override this
 */
void __weak crash_map_reserved_pages(void)
{}

void __weak crash_unmap_reserved_pages(void)
{}

1260
#ifdef CONFIG_COMPAT
1261 1262 1263 1264
COMPAT_SYSCALL_DEFINE4(kexec_load, compat_ulong_t, entry,
		       compat_ulong_t, nr_segments,
		       struct compat_kexec_segment __user *, segments,
		       compat_ulong_t, flags)
1265 1266 1267 1268 1269 1270 1271 1272
{
	struct compat_kexec_segment in;
	struct kexec_segment out, __user *ksegments;
	unsigned long i, result;

	/* Don't allow clients that don't understand the native
	 * architecture to do anything.
	 */
M
Maneesh Soni 已提交
1273
	if ((flags & KEXEC_ARCH_MASK) == KEXEC_ARCH_DEFAULT)
1274 1275
		return -EINVAL;

M
Maneesh Soni 已提交
1276
	if (nr_segments > KEXEC_SEGMENT_MAX)
1277 1278 1279
		return -EINVAL;

	ksegments = compat_alloc_user_space(nr_segments * sizeof(out));
1280
	for (i = 0; i < nr_segments; i++) {
1281
		result = copy_from_user(&in, &segments[i], sizeof(in));
M
Maneesh Soni 已提交
1282
		if (result)
1283 1284 1285 1286 1287 1288 1289 1290
			return -EFAULT;

		out.buf   = compat_ptr(in.buf);
		out.bufsz = in.bufsz;
		out.mem   = in.mem;
		out.memsz = in.memsz;

		result = copy_to_user(&ksegments[i], &out, sizeof(out));
M
Maneesh Soni 已提交
1291
		if (result)
1292 1293 1294 1295 1296 1297 1298
			return -EFAULT;
	}

	return sys_kexec_load(entry, nr_segments, ksegments, flags);
}
#endif

1299 1300 1301 1302
SYSCALL_DEFINE5(kexec_file_load, int, kernel_fd, int, initrd_fd,
		unsigned long, cmdline_len, const char __user *, cmdline_ptr,
		unsigned long, flags)
{
1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368
	int ret = 0, i;
	struct kimage **dest_image, *image;

	/* We only trust the superuser with rebooting the system. */
	if (!capable(CAP_SYS_BOOT) || kexec_load_disabled)
		return -EPERM;

	/* Make sure we have a legal set of flags */
	if (flags != (flags & KEXEC_FILE_FLAGS))
		return -EINVAL;

	image = NULL;

	if (!mutex_trylock(&kexec_mutex))
		return -EBUSY;

	dest_image = &kexec_image;
	if (flags & KEXEC_FILE_ON_CRASH)
		dest_image = &kexec_crash_image;

	if (flags & KEXEC_FILE_UNLOAD)
		goto exchange;

	/*
	 * In case of crash, new kernel gets loaded in reserved region. It is
	 * same memory where old crash kernel might be loaded. Free any
	 * current crash dump kernel before we corrupt it.
	 */
	if (flags & KEXEC_FILE_ON_CRASH)
		kimage_free(xchg(&kexec_crash_image, NULL));

	ret = kimage_file_alloc_init(&image, kernel_fd, initrd_fd, cmdline_ptr,
				     cmdline_len, flags);
	if (ret)
		goto out;

	ret = machine_kexec_prepare(image);
	if (ret)
		goto out;

	for (i = 0; i < image->nr_segments; i++) {
		struct kexec_segment *ksegment;

		ksegment = &image->segment[i];
		pr_debug("Loading segment %d: buf=0x%p bufsz=0x%zx mem=0x%lx memsz=0x%zx\n",
			 i, ksegment->buf, ksegment->bufsz, ksegment->mem,
			 ksegment->memsz);

		ret = kimage_load_segment(image, &image->segment[i]);
		if (ret)
			goto out;
	}

	kimage_terminate(image);

	/*
	 * Free up any temporary buffers allocated which are not needed
	 * after image has been loaded
	 */
	kimage_file_post_load_cleanup(image);
exchange:
	image = xchg(dest_image, image);
out:
	mutex_unlock(&kexec_mutex);
	kimage_free(image);
	return ret;
1369 1370
}

1371
void crash_kexec(struct pt_regs *regs)
1372
{
1373
	/* Take the kexec_mutex here to prevent sys_kexec_load
1374 1375 1376 1377 1378 1379 1380
	 * running on one cpu from replacing the crash kernel
	 * we are using after a panic on a different cpu.
	 *
	 * If the crash kernel was not located in a fixed area
	 * of memory the xchg(&kexec_crash_image) would be
	 * sufficient.  But since I reuse the memory...
	 */
1381
	if (mutex_trylock(&kexec_mutex)) {
1382
		if (kexec_crash_image) {
1383
			struct pt_regs fixed_regs;
1384

1385
			crash_setup_regs(&fixed_regs, regs);
K
Ken'ichi Ohmichi 已提交
1386
			crash_save_vmcoreinfo();
1387
			machine_crash_shutdown(&fixed_regs);
1388
			machine_kexec(kexec_crash_image);
1389
		}
1390
		mutex_unlock(&kexec_mutex);
1391 1392
	}
}
1393

1394 1395
size_t crash_get_memory_size(void)
{
1396
	size_t size = 0;
1397
	mutex_lock(&kexec_mutex);
1398
	if (crashk_res.end != crashk_res.start)
1399
		size = resource_size(&crashk_res);
1400 1401 1402 1403
	mutex_unlock(&kexec_mutex);
	return size;
}

1404 1405
void __weak crash_free_reserved_phys_range(unsigned long begin,
					   unsigned long end)
1406 1407 1408
{
	unsigned long addr;

1409 1410
	for (addr = begin; addr < end; addr += PAGE_SIZE)
		free_reserved_page(pfn_to_page(addr >> PAGE_SHIFT));
1411 1412 1413 1414 1415 1416
}

int crash_shrink_memory(unsigned long new_size)
{
	int ret = 0;
	unsigned long start, end;
1417
	unsigned long old_size;
1418
	struct resource *ram_res;
1419 1420 1421 1422 1423 1424 1425 1426 1427

	mutex_lock(&kexec_mutex);

	if (kexec_crash_image) {
		ret = -ENOENT;
		goto unlock;
	}
	start = crashk_res.start;
	end = crashk_res.end;
1428 1429 1430
	old_size = (end == 0) ? 0 : end - start + 1;
	if (new_size >= old_size) {
		ret = (new_size == old_size) ? 0 : -EINVAL;
1431 1432 1433
		goto unlock;
	}

1434 1435 1436 1437 1438 1439
	ram_res = kzalloc(sizeof(*ram_res), GFP_KERNEL);
	if (!ram_res) {
		ret = -ENOMEM;
		goto unlock;
	}

1440 1441
	start = roundup(start, KEXEC_CRASH_MEM_ALIGN);
	end = roundup(start + new_size, KEXEC_CRASH_MEM_ALIGN);
1442

1443
	crash_map_reserved_pages();
1444
	crash_free_reserved_phys_range(end, crashk_res.end);
1445

1446
	if ((start == end) && (crashk_res.parent != NULL))
1447
		release_resource(&crashk_res);
1448 1449 1450 1451 1452 1453

	ram_res->start = end;
	ram_res->end = crashk_res.end;
	ram_res->flags = IORESOURCE_BUSY | IORESOURCE_MEM;
	ram_res->name = "System RAM";

1454
	crashk_res.end = end - 1;
1455 1456

	insert_resource(&iomem_resource, ram_res);
1457
	crash_unmap_reserved_pages();
1458 1459 1460 1461 1462 1463

unlock:
	mutex_unlock(&kexec_mutex);
	return ret;
}

1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496
static u32 *append_elf_note(u32 *buf, char *name, unsigned type, void *data,
			    size_t data_len)
{
	struct elf_note note;

	note.n_namesz = strlen(name) + 1;
	note.n_descsz = data_len;
	note.n_type   = type;
	memcpy(buf, &note, sizeof(note));
	buf += (sizeof(note) + 3)/4;
	memcpy(buf, name, note.n_namesz);
	buf += (note.n_namesz + 3)/4;
	memcpy(buf, data, note.n_descsz);
	buf += (note.n_descsz + 3)/4;

	return buf;
}

static void final_note(u32 *buf)
{
	struct elf_note note;

	note.n_namesz = 0;
	note.n_descsz = 0;
	note.n_type   = 0;
	memcpy(buf, &note, sizeof(note));
}

void crash_save_cpu(struct pt_regs *regs, int cpu)
{
	struct elf_prstatus prstatus;
	u32 *buf;

1497
	if ((cpu < 0) || (cpu >= nr_cpu_ids))
1498 1499 1500 1501 1502 1503 1504 1505 1506
		return;

	/* Using ELF notes here is opportunistic.
	 * I need a well defined structure format
	 * for the data I pass, and I need tags
	 * on the data to indicate what information I have
	 * squirrelled away.  ELF notes happen to provide
	 * all of that, so there is no need to invent something new.
	 */
1507
	buf = (u32 *)per_cpu_ptr(crash_notes, cpu);
1508 1509 1510 1511
	if (!buf)
		return;
	memset(&prstatus, 0, sizeof(prstatus));
	prstatus.pr_pid = current->pid;
T
Tejun Heo 已提交
1512
	elf_core_copy_kernel_regs(&prstatus.pr_reg, regs);
1513
	buf = append_elf_note(buf, KEXEC_CORE_NOTE_NAME, NT_PRSTATUS,
1514
			      &prstatus, sizeof(prstatus));
1515 1516 1517
	final_note(buf);
}

1518 1519 1520 1521 1522
static int __init crash_notes_memory_init(void)
{
	/* Allocate memory for saving cpu registers. */
	crash_notes = alloc_percpu(note_buf_t);
	if (!crash_notes) {
1523
		pr_warn("Kexec: Memory allocation for saving cpu register states failed\n");
1524 1525 1526 1527
		return -ENOMEM;
	}
	return 0;
}
1528
subsys_initcall(crash_notes_memory_init);
K
Ken'ichi Ohmichi 已提交
1529

B
Bernhard Walle 已提交
1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544

/*
 * parsing the "crashkernel" commandline
 *
 * this code is intended to be called from architecture specific code
 */


/*
 * This function parses command lines in the format
 *
 *   crashkernel=ramsize-range:size[,...][@offset]
 *
 * The function returns 0 on success and -EINVAL on failure.
 */
1545 1546 1547 1548
static int __init parse_crashkernel_mem(char *cmdline,
					unsigned long long system_ram,
					unsigned long long *crash_size,
					unsigned long long *crash_base)
B
Bernhard Walle 已提交
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
{
	char *cur = cmdline, *tmp;

	/* for each entry of the comma-separated list */
	do {
		unsigned long long start, end = ULLONG_MAX, size;

		/* get the start of the range */
		start = memparse(cur, &tmp);
		if (cur == tmp) {
1559
			pr_warn("crashkernel: Memory value expected\n");
B
Bernhard Walle 已提交
1560 1561 1562 1563
			return -EINVAL;
		}
		cur = tmp;
		if (*cur != '-') {
1564
			pr_warn("crashkernel: '-' expected\n");
B
Bernhard Walle 已提交
1565 1566 1567 1568 1569 1570 1571 1572
			return -EINVAL;
		}
		cur++;

		/* if no ':' is here, than we read the end */
		if (*cur != ':') {
			end = memparse(cur, &tmp);
			if (cur == tmp) {
1573
				pr_warn("crashkernel: Memory value expected\n");
B
Bernhard Walle 已提交
1574 1575 1576 1577
				return -EINVAL;
			}
			cur = tmp;
			if (end <= start) {
1578
				pr_warn("crashkernel: end <= start\n");
B
Bernhard Walle 已提交
1579 1580 1581 1582 1583
				return -EINVAL;
			}
		}

		if (*cur != ':') {
1584
			pr_warn("crashkernel: ':' expected\n");
B
Bernhard Walle 已提交
1585 1586 1587 1588 1589 1590
			return -EINVAL;
		}
		cur++;

		size = memparse(cur, &tmp);
		if (cur == tmp) {
1591
			pr_warn("Memory value expected\n");
B
Bernhard Walle 已提交
1592 1593 1594 1595
			return -EINVAL;
		}
		cur = tmp;
		if (size >= system_ram) {
1596
			pr_warn("crashkernel: invalid size\n");
B
Bernhard Walle 已提交
1597 1598 1599 1600
			return -EINVAL;
		}

		/* match ? */
1601
		if (system_ram >= start && system_ram < end) {
B
Bernhard Walle 已提交
1602 1603 1604 1605 1606 1607
			*crash_size = size;
			break;
		}
	} while (*cur++ == ',');

	if (*crash_size > 0) {
1608
		while (*cur && *cur != ' ' && *cur != '@')
B
Bernhard Walle 已提交
1609 1610 1611 1612 1613
			cur++;
		if (*cur == '@') {
			cur++;
			*crash_base = memparse(cur, &tmp);
			if (cur == tmp) {
1614
				pr_warn("Memory value expected after '@'\n");
B
Bernhard Walle 已提交
1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625
				return -EINVAL;
			}
		}
	}

	return 0;
}

/*
 * That function parses "simple" (old) crashkernel command lines like
 *
1626
 *	crashkernel=size[@offset]
B
Bernhard Walle 已提交
1627 1628 1629
 *
 * It returns 0 on success and -EINVAL on failure.
 */
1630 1631 1632
static int __init parse_crashkernel_simple(char *cmdline,
					   unsigned long long *crash_size,
					   unsigned long long *crash_base)
B
Bernhard Walle 已提交
1633 1634 1635 1636 1637
{
	char *cur = cmdline;

	*crash_size = memparse(cmdline, &cur);
	if (cmdline == cur) {
1638
		pr_warn("crashkernel: memory value expected\n");
B
Bernhard Walle 已提交
1639 1640 1641 1642 1643
		return -EINVAL;
	}

	if (*cur == '@')
		*crash_base = memparse(cur+1, &cur);
1644
	else if (*cur != ' ' && *cur != '\0') {
1645
		pr_warn("crashkernel: unrecognized char\n");
1646 1647
		return -EINVAL;
	}
B
Bernhard Walle 已提交
1648 1649 1650 1651

	return 0;
}

1652 1653 1654 1655 1656 1657 1658 1659 1660
#define SUFFIX_HIGH 0
#define SUFFIX_LOW  1
#define SUFFIX_NULL 2
static __initdata char *suffix_tbl[] = {
	[SUFFIX_HIGH] = ",high",
	[SUFFIX_LOW]  = ",low",
	[SUFFIX_NULL] = NULL,
};

B
Bernhard Walle 已提交
1661
/*
1662 1663 1664 1665 1666
 * That function parses "suffix"  crashkernel command lines like
 *
 *	crashkernel=size,[high|low]
 *
 * It returns 0 on success and -EINVAL on failure.
B
Bernhard Walle 已提交
1667
 */
1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735
static int __init parse_crashkernel_suffix(char *cmdline,
					   unsigned long long	*crash_size,
					   unsigned long long	*crash_base,
					   const char *suffix)
{
	char *cur = cmdline;

	*crash_size = memparse(cmdline, &cur);
	if (cmdline == cur) {
		pr_warn("crashkernel: memory value expected\n");
		return -EINVAL;
	}

	/* check with suffix */
	if (strncmp(cur, suffix, strlen(suffix))) {
		pr_warn("crashkernel: unrecognized char\n");
		return -EINVAL;
	}
	cur += strlen(suffix);
	if (*cur != ' ' && *cur != '\0') {
		pr_warn("crashkernel: unrecognized char\n");
		return -EINVAL;
	}

	return 0;
}

static __init char *get_last_crashkernel(char *cmdline,
			     const char *name,
			     const char *suffix)
{
	char *p = cmdline, *ck_cmdline = NULL;

	/* find crashkernel and use the last one if there are more */
	p = strstr(p, name);
	while (p) {
		char *end_p = strchr(p, ' ');
		char *q;

		if (!end_p)
			end_p = p + strlen(p);

		if (!suffix) {
			int i;

			/* skip the one with any known suffix */
			for (i = 0; suffix_tbl[i]; i++) {
				q = end_p - strlen(suffix_tbl[i]);
				if (!strncmp(q, suffix_tbl[i],
					     strlen(suffix_tbl[i])))
					goto next;
			}
			ck_cmdline = p;
		} else {
			q = end_p - strlen(suffix);
			if (!strncmp(q, suffix, strlen(suffix)))
				ck_cmdline = p;
		}
next:
		p = strstr(p+1, name);
	}

	if (!ck_cmdline)
		return NULL;

	return ck_cmdline;
}

1736
static int __init __parse_crashkernel(char *cmdline,
B
Bernhard Walle 已提交
1737 1738
			     unsigned long long system_ram,
			     unsigned long long *crash_size,
1739
			     unsigned long long *crash_base,
1740 1741
			     const char *name,
			     const char *suffix)
B
Bernhard Walle 已提交
1742 1743
{
	char	*first_colon, *first_space;
1744
	char	*ck_cmdline;
B
Bernhard Walle 已提交
1745 1746 1747 1748 1749

	BUG_ON(!crash_size || !crash_base);
	*crash_size = 0;
	*crash_base = 0;

1750
	ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
B
Bernhard Walle 已提交
1751 1752 1753 1754

	if (!ck_cmdline)
		return -EINVAL;

1755
	ck_cmdline += strlen(name);
B
Bernhard Walle 已提交
1756

1757 1758 1759
	if (suffix)
		return parse_crashkernel_suffix(ck_cmdline, crash_size,
				crash_base, suffix);
B
Bernhard Walle 已提交
1760 1761 1762 1763 1764 1765 1766 1767 1768 1769
	/*
	 * if the commandline contains a ':', then that's the extended
	 * syntax -- if not, it must be the classic syntax
	 */
	first_colon = strchr(ck_cmdline, ':');
	first_space = strchr(ck_cmdline, ' ');
	if (first_colon && (!first_space || first_colon < first_space))
		return parse_crashkernel_mem(ck_cmdline, system_ram,
				crash_size, crash_base);

X
Xishi Qiu 已提交
1770
	return parse_crashkernel_simple(ck_cmdline, crash_size, crash_base);
B
Bernhard Walle 已提交
1771 1772
}

1773 1774 1775 1776
/*
 * That function is the entry point for command line parsing and should be
 * called from the arch-specific code.
 */
1777 1778 1779 1780 1781 1782
int __init parse_crashkernel(char *cmdline,
			     unsigned long long system_ram,
			     unsigned long long *crash_size,
			     unsigned long long *crash_base)
{
	return __parse_crashkernel(cmdline, system_ram, crash_size, crash_base,
1783
					"crashkernel=", NULL);
1784
}
1785 1786 1787 1788 1789 1790 1791

int __init parse_crashkernel_high(char *cmdline,
			     unsigned long long system_ram,
			     unsigned long long *crash_size,
			     unsigned long long *crash_base)
{
	return __parse_crashkernel(cmdline, system_ram, crash_size, crash_base,
1792
				"crashkernel=", suffix_tbl[SUFFIX_HIGH]);
1793
}
1794 1795 1796 1797 1798 1799 1800

int __init parse_crashkernel_low(char *cmdline,
			     unsigned long long system_ram,
			     unsigned long long *crash_size,
			     unsigned long long *crash_base)
{
	return __parse_crashkernel(cmdline, system_ram, crash_size, crash_base,
1801
				"crashkernel=", suffix_tbl[SUFFIX_LOW]);
1802
}
B
Bernhard Walle 已提交
1803

1804
static void update_vmcoreinfo_note(void)
K
Ken'ichi Ohmichi 已提交
1805
{
1806
	u32 *buf = vmcoreinfo_note;
K
Ken'ichi Ohmichi 已提交
1807 1808 1809 1810 1811 1812 1813 1814

	if (!vmcoreinfo_size)
		return;
	buf = append_elf_note(buf, VMCOREINFO_NOTE_NAME, 0, vmcoreinfo_data,
			      vmcoreinfo_size);
	final_note(buf);
}

1815 1816
void crash_save_vmcoreinfo(void)
{
1817
	vmcoreinfo_append_str("CRASHTIME=%ld\n", get_seconds());
1818 1819 1820
	update_vmcoreinfo_note();
}

K
Ken'ichi Ohmichi 已提交
1821 1822 1823 1824
void vmcoreinfo_append_str(const char *fmt, ...)
{
	va_list args;
	char buf[0x50];
1825
	size_t r;
K
Ken'ichi Ohmichi 已提交
1826 1827

	va_start(args, fmt);
1828
	r = vscnprintf(buf, sizeof(buf), fmt, args);
K
Ken'ichi Ohmichi 已提交
1829 1830
	va_end(args);

1831
	r = min(r, vmcoreinfo_max_size - vmcoreinfo_size);
K
Ken'ichi Ohmichi 已提交
1832 1833 1834 1835 1836 1837 1838 1839 1840 1841

	memcpy(&vmcoreinfo_data[vmcoreinfo_size], buf, r);

	vmcoreinfo_size += r;
}

/*
 * provide an empty default implementation here -- architecture
 * code may override this
 */
1842
void __weak arch_crash_save_vmcoreinfo(void)
K
Ken'ichi Ohmichi 已提交
1843 1844
{}

1845
unsigned long __weak paddr_vmcoreinfo_note(void)
K
Ken'ichi Ohmichi 已提交
1846 1847 1848 1849 1850 1851
{
	return __pa((unsigned long)(char *)&vmcoreinfo_note);
}

static int __init crash_save_vmcoreinfo_init(void)
{
1852 1853
	VMCOREINFO_OSRELEASE(init_uts_ns.name.release);
	VMCOREINFO_PAGESIZE(PAGE_SIZE);
K
Ken'ichi Ohmichi 已提交
1854

1855 1856
	VMCOREINFO_SYMBOL(init_uts_ns);
	VMCOREINFO_SYMBOL(node_online_map);
1857
#ifdef CONFIG_MMU
1858
	VMCOREINFO_SYMBOL(swapper_pg_dir);
1859
#endif
1860
	VMCOREINFO_SYMBOL(_stext);
1861
	VMCOREINFO_SYMBOL(vmap_area_list);
K
Ken'ichi Ohmichi 已提交
1862 1863

#ifndef CONFIG_NEED_MULTIPLE_NODES
1864 1865
	VMCOREINFO_SYMBOL(mem_map);
	VMCOREINFO_SYMBOL(contig_page_data);
K
Ken'ichi Ohmichi 已提交
1866 1867
#endif
#ifdef CONFIG_SPARSEMEM
1868 1869
	VMCOREINFO_SYMBOL(mem_section);
	VMCOREINFO_LENGTH(mem_section, NR_SECTION_ROOTS);
1870
	VMCOREINFO_STRUCT_SIZE(mem_section);
1871
	VMCOREINFO_OFFSET(mem_section, section_mem_map);
K
Ken'ichi Ohmichi 已提交
1872
#endif
1873 1874 1875 1876 1877 1878
	VMCOREINFO_STRUCT_SIZE(page);
	VMCOREINFO_STRUCT_SIZE(pglist_data);
	VMCOREINFO_STRUCT_SIZE(zone);
	VMCOREINFO_STRUCT_SIZE(free_area);
	VMCOREINFO_STRUCT_SIZE(list_head);
	VMCOREINFO_SIZE(nodemask_t);
1879 1880 1881 1882
	VMCOREINFO_OFFSET(page, flags);
	VMCOREINFO_OFFSET(page, _count);
	VMCOREINFO_OFFSET(page, mapping);
	VMCOREINFO_OFFSET(page, lru);
1883 1884
	VMCOREINFO_OFFSET(page, _mapcount);
	VMCOREINFO_OFFSET(page, private);
1885 1886
	VMCOREINFO_OFFSET(pglist_data, node_zones);
	VMCOREINFO_OFFSET(pglist_data, nr_zones);
K
Ken'ichi Ohmichi 已提交
1887
#ifdef CONFIG_FLAT_NODE_MEM_MAP
1888
	VMCOREINFO_OFFSET(pglist_data, node_mem_map);
K
Ken'ichi Ohmichi 已提交
1889
#endif
1890 1891 1892 1893 1894 1895 1896 1897 1898
	VMCOREINFO_OFFSET(pglist_data, node_start_pfn);
	VMCOREINFO_OFFSET(pglist_data, node_spanned_pages);
	VMCOREINFO_OFFSET(pglist_data, node_id);
	VMCOREINFO_OFFSET(zone, free_area);
	VMCOREINFO_OFFSET(zone, vm_stat);
	VMCOREINFO_OFFSET(zone, spanned_pages);
	VMCOREINFO_OFFSET(free_area, free_list);
	VMCOREINFO_OFFSET(list_head, next);
	VMCOREINFO_OFFSET(list_head, prev);
1899 1900
	VMCOREINFO_OFFSET(vmap_area, va_start);
	VMCOREINFO_OFFSET(vmap_area, list);
1901
	VMCOREINFO_LENGTH(zone.free_area, MAX_ORDER);
1902
	log_buf_kexec_setup();
1903
	VMCOREINFO_LENGTH(free_area.free_list, MIGRATE_TYPES);
1904
	VMCOREINFO_NUMBER(NR_FREE_PAGES);
1905 1906 1907
	VMCOREINFO_NUMBER(PG_lru);
	VMCOREINFO_NUMBER(PG_private);
	VMCOREINFO_NUMBER(PG_swapcache);
1908
	VMCOREINFO_NUMBER(PG_slab);
1909 1910 1911
#ifdef CONFIG_MEMORY_FAILURE
	VMCOREINFO_NUMBER(PG_hwpoison);
#endif
1912
	VMCOREINFO_NUMBER(PG_head_mask);
1913
	VMCOREINFO_NUMBER(PAGE_BUDDY_MAPCOUNT_VALUE);
1914
#ifdef CONFIG_HUGETLBFS
1915
	VMCOREINFO_SYMBOL(free_huge_page);
1916
#endif
K
Ken'ichi Ohmichi 已提交
1917 1918

	arch_crash_save_vmcoreinfo();
1919
	update_vmcoreinfo_note();
K
Ken'ichi Ohmichi 已提交
1920 1921 1922 1923

	return 0;
}

1924
subsys_initcall(crash_save_vmcoreinfo_init);
H
Huang Ying 已提交
1925

1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095
static int __kexec_add_segment(struct kimage *image, char *buf,
			       unsigned long bufsz, unsigned long mem,
			       unsigned long memsz)
{
	struct kexec_segment *ksegment;

	ksegment = &image->segment[image->nr_segments];
	ksegment->kbuf = buf;
	ksegment->bufsz = bufsz;
	ksegment->mem = mem;
	ksegment->memsz = memsz;
	image->nr_segments++;

	return 0;
}

static int locate_mem_hole_top_down(unsigned long start, unsigned long end,
				    struct kexec_buf *kbuf)
{
	struct kimage *image = kbuf->image;
	unsigned long temp_start, temp_end;

	temp_end = min(end, kbuf->buf_max);
	temp_start = temp_end - kbuf->memsz;

	do {
		/* align down start */
		temp_start = temp_start & (~(kbuf->buf_align - 1));

		if (temp_start < start || temp_start < kbuf->buf_min)
			return 0;

		temp_end = temp_start + kbuf->memsz - 1;

		/*
		 * Make sure this does not conflict with any of existing
		 * segments
		 */
		if (kimage_is_destination_range(image, temp_start, temp_end)) {
			temp_start = temp_start - PAGE_SIZE;
			continue;
		}

		/* We found a suitable memory range */
		break;
	} while (1);

	/* If we are here, we found a suitable memory range */
	__kexec_add_segment(image, kbuf->buffer, kbuf->bufsz, temp_start,
			    kbuf->memsz);

	/* Success, stop navigating through remaining System RAM ranges */
	return 1;
}

static int locate_mem_hole_bottom_up(unsigned long start, unsigned long end,
				     struct kexec_buf *kbuf)
{
	struct kimage *image = kbuf->image;
	unsigned long temp_start, temp_end;

	temp_start = max(start, kbuf->buf_min);

	do {
		temp_start = ALIGN(temp_start, kbuf->buf_align);
		temp_end = temp_start + kbuf->memsz - 1;

		if (temp_end > end || temp_end > kbuf->buf_max)
			return 0;
		/*
		 * Make sure this does not conflict with any of existing
		 * segments
		 */
		if (kimage_is_destination_range(image, temp_start, temp_end)) {
			temp_start = temp_start + PAGE_SIZE;
			continue;
		}

		/* We found a suitable memory range */
		break;
	} while (1);

	/* If we are here, we found a suitable memory range */
	__kexec_add_segment(image, kbuf->buffer, kbuf->bufsz, temp_start,
			    kbuf->memsz);

	/* Success, stop navigating through remaining System RAM ranges */
	return 1;
}

static int locate_mem_hole_callback(u64 start, u64 end, void *arg)
{
	struct kexec_buf *kbuf = (struct kexec_buf *)arg;
	unsigned long sz = end - start + 1;

	/* Returning 0 will take to next memory range */
	if (sz < kbuf->memsz)
		return 0;

	if (end < kbuf->buf_min || start > kbuf->buf_max)
		return 0;

	/*
	 * Allocate memory top down with-in ram range. Otherwise bottom up
	 * allocation.
	 */
	if (kbuf->top_down)
		return locate_mem_hole_top_down(start, end, kbuf);
	return locate_mem_hole_bottom_up(start, end, kbuf);
}

/*
 * Helper function for placing a buffer in a kexec segment. This assumes
 * that kexec_mutex is held.
 */
int kexec_add_buffer(struct kimage *image, char *buffer, unsigned long bufsz,
		     unsigned long memsz, unsigned long buf_align,
		     unsigned long buf_min, unsigned long buf_max,
		     bool top_down, unsigned long *load_addr)
{

	struct kexec_segment *ksegment;
	struct kexec_buf buf, *kbuf;
	int ret;

	/* Currently adding segment this way is allowed only in file mode */
	if (!image->file_mode)
		return -EINVAL;

	if (image->nr_segments >= KEXEC_SEGMENT_MAX)
		return -EINVAL;

	/*
	 * Make sure we are not trying to add buffer after allocating
	 * control pages. All segments need to be placed first before
	 * any control pages are allocated. As control page allocation
	 * logic goes through list of segments to make sure there are
	 * no destination overlaps.
	 */
	if (!list_empty(&image->control_pages)) {
		WARN_ON(1);
		return -EINVAL;
	}

	memset(&buf, 0, sizeof(struct kexec_buf));
	kbuf = &buf;
	kbuf->image = image;
	kbuf->buffer = buffer;
	kbuf->bufsz = bufsz;

	kbuf->memsz = ALIGN(memsz, PAGE_SIZE);
	kbuf->buf_align = max(buf_align, PAGE_SIZE);
	kbuf->buf_min = buf_min;
	kbuf->buf_max = buf_max;
	kbuf->top_down = top_down;

	/* Walk the RAM ranges and allocate a suitable range for the buffer */
	ret = walk_system_ram_res(0, -1, kbuf, locate_mem_hole_callback);
	if (ret != 1) {
		/* A suitable memory range could not be found for buffer */
		return -EADDRNOTAVAIL;
	}

	/* Found a suitable memory range */
	ksegment = &image->segment[image->nr_segments - 1];
	*load_addr = ksegment->mem;
	return 0;
}


2096 2097 2098
/*
 * Move into place and start executing a preloaded standalone
 * executable.  If nothing was preloaded return an error.
H
Huang Ying 已提交
2099 2100 2101 2102 2103
 */
int kernel_kexec(void)
{
	int error = 0;

2104
	if (!mutex_trylock(&kexec_mutex))
H
Huang Ying 已提交
2105 2106 2107 2108 2109 2110 2111
		return -EBUSY;
	if (!kexec_image) {
		error = -EINVAL;
		goto Unlock;
	}

#ifdef CONFIG_KEXEC_JUMP
2112
	if (kexec_image->preserve_context) {
2113
		lock_system_sleep();
2114 2115 2116 2117 2118 2119 2120
		pm_prepare_console();
		error = freeze_processes();
		if (error) {
			error = -EBUSY;
			goto Restore_console;
		}
		suspend_console();
2121
		error = dpm_suspend_start(PMSG_FREEZE);
2122 2123
		if (error)
			goto Resume_console;
2124
		/* At this point, dpm_suspend_start() has been called,
2125 2126
		 * but *not* dpm_suspend_end(). We *must* call
		 * dpm_suspend_end() now.  Otherwise, drivers for
2127 2128 2129 2130
		 * some devices (e.g. interrupt controllers) become
		 * desynchronized with the actual state of the
		 * hardware at resume time, and evil weirdness ensues.
		 */
2131
		error = dpm_suspend_end(PMSG_FREEZE);
2132
		if (error)
2133 2134 2135 2136
			goto Resume_devices;
		error = disable_nonboot_cpus();
		if (error)
			goto Enable_cpus;
2137
		local_irq_disable();
2138
		error = syscore_suspend();
2139
		if (error)
2140
			goto Enable_irqs;
2141
	} else
H
Huang Ying 已提交
2142
#endif
2143
	{
2144
		kexec_in_progress = true;
2145
		kernel_restart_prepare(NULL);
V
Vivek Goyal 已提交
2146
		migrate_to_reboot_cpu();
2147 2148 2149 2150 2151 2152 2153 2154

		/*
		 * migrate_to_reboot_cpu() disables CPU hotplug assuming that
		 * no further code needs to use CPU hotplug (which is true in
		 * the reboot case). However, the kexec path depends on using
		 * CPU hotplug again; so re-enable it here.
		 */
		cpu_hotplug_enable();
2155
		pr_emerg("Starting new kernel\n");
H
Huang Ying 已提交
2156 2157 2158 2159 2160 2161
		machine_shutdown();
	}

	machine_kexec(kexec_image);

#ifdef CONFIG_KEXEC_JUMP
2162
	if (kexec_image->preserve_context) {
2163
		syscore_resume();
2164
 Enable_irqs:
H
Huang Ying 已提交
2165
		local_irq_enable();
2166
 Enable_cpus:
2167
		enable_nonboot_cpus();
2168
		dpm_resume_start(PMSG_RESTORE);
2169
 Resume_devices:
2170
		dpm_resume_end(PMSG_RESTORE);
2171 2172 2173 2174 2175
 Resume_console:
		resume_console();
		thaw_processes();
 Restore_console:
		pm_restore_console();
2176
		unlock_system_sleep();
H
Huang Ying 已提交
2177
	}
2178
#endif
H
Huang Ying 已提交
2179 2180

 Unlock:
2181
	mutex_unlock(&kexec_mutex);
H
Huang Ying 已提交
2182 2183
	return error;
}