kexec.c 42.5 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
#include <linux/capability.h>
10 11 12 13 14
#include <linux/mm.h>
#include <linux/file.h>
#include <linux/slab.h>
#include <linux/fs.h>
#include <linux/kexec.h>
15
#include <linux/mutex.h>
16 17 18 19 20
#include <linux/list.h>
#include <linux/highmem.h>
#include <linux/syscalls.h>
#include <linux/reboot.h>
#include <linux/ioport.h>
21
#include <linux/hardirq.h>
22 23
#include <linux/elf.h>
#include <linux/elfcore.h>
K
Ken'ichi Ohmichi 已提交
24 25
#include <linux/utsname.h>
#include <linux/numa.h>
H
Huang Ying 已提交
26 27
#include <linux/suspend.h>
#include <linux/device.h>
28 29 30 31
#include <linux/freezer.h>
#include <linux/pm.h>
#include <linux/cpu.h>
#include <linux/console.h>
32
#include <linux/vmalloc.h>
33
#include <linux/swap.h>
34
#include <linux/syscore_ops.h>
35

36 37 38
#include <asm/page.h>
#include <asm/uaccess.h>
#include <asm/io.h>
K
Ken'ichi Ohmichi 已提交
39
#include <asm/sections.h>
40

41
/* Per cpu memory for storing cpu states in case of system crash. */
42
note_buf_t __percpu *crash_notes;
43

K
Ken'ichi Ohmichi 已提交
44
/* vmcoreinfo stuff */
45
static unsigned char vmcoreinfo_data[VMCOREINFO_BYTES];
K
Ken'ichi Ohmichi 已提交
46
u32 vmcoreinfo_note[VMCOREINFO_NOTE_SIZE/4];
47 48
size_t vmcoreinfo_size;
size_t vmcoreinfo_max_size = sizeof(vmcoreinfo_data);
K
Ken'ichi Ohmichi 已提交
49

50 51 52
/* Flag to indicate we are going to kexec a new kernel */
bool kexec_in_progress = false;

53 54 55 56 57 58 59
/* 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
};
60
struct resource crashk_low_res = {
61
	.name  = "Crash kernel",
62 63 64 65
	.start = 0,
	.end   = 0,
	.flags = IORESOURCE_BUSY | IORESOURCE_MEM
};
66

67 68
int kexec_should_crash(struct task_struct *p)
{
69
	if (in_interrupt() || !p->pid || is_global_init(p) || panic_on_oops)
70 71 72 73
		return 1;
	return 0;
}

74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89
/*
 * 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
90
 * is given by KEXEC_CONTROL_PAGE_SIZE.  In the best case only a single
91 92 93 94 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
 * 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
120 121 122
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
123
				       gfp_t gfp_mask,
M
Maneesh Soni 已提交
124
				       unsigned long dest);
125 126

static int do_kimage_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
127 128
	                    unsigned long nr_segments,
                            struct kexec_segment __user *segments)
129 130 131 132 133 134 135 136
{
	size_t segment_bytes;
	struct kimage *image;
	unsigned long i;
	int result;

	/* Allocate a controlling structure */
	result = -ENOMEM;
137
	image = kzalloc(sizeof(*image), GFP_KERNEL);
M
Maneesh Soni 已提交
138
	if (!image)
139
		goto out;
M
Maneesh Soni 已提交
140

141 142 143 144 145 146 147 148 149 150 151 152 153
	image->head = 0;
	image->entry = &image->head;
	image->last_entry = &image->head;
	image->control_page = ~0; /* By default this does not apply */
	image->start = entry;
	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);

L
Lucas De Marchi 已提交
154
	/* Initialize the list of unusable pages */
155 156 157 158 159 160
	INIT_LIST_HEAD(&image->unuseable_pages);

	/* Read in the segments */
	image->nr_segments = nr_segments;
	segment_bytes = nr_segments * sizeof(*segments);
	result = copy_from_user(image->segment, segments, segment_bytes);
161 162
	if (result) {
		result = -EFAULT;
163
		goto out;
164
	}
165 166 167 168 169 170 171 172

	/*
	 * 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
173
	 * the destination addresses are page aligned.  Too many
174 175 176 177 178 179 180 181
	 * 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
182

183 184 185 186 187 188 189 190 191 192 193 194 195 196
		mstart = image->segment[i].mem;
		mend   = mstart + image->segment[i].memsz;
		if ((mstart & ~PAGE_MASK) || (mend & ~PAGE_MASK))
			goto out;
		if (mend >= KEXEC_DESTINATION_MEMORY_LIMIT)
			goto out;
	}

	/* 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
197
	for (i = 0; i < nr_segments; i++) {
198 199
		unsigned long mstart, mend;
		unsigned long j;
M
Maneesh Soni 已提交
200

201 202
		mstart = image->segment[i].mem;
		mend   = mstart + image->segment[i].memsz;
M
Maneesh Soni 已提交
203
		for (j = 0; j < i; j++) {
204 205 206 207 208 209 210 211 212 213 214 215 216 217 218
			unsigned long pstart, pend;
			pstart = image->segment[j].mem;
			pend   = pstart + image->segment[j].memsz;
			/* Do the segments overlap ? */
			if ((mend > pstart) && (mstart < pend))
				goto out;
		}
	}

	/* 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
219
	for (i = 0; i < nr_segments; i++) {
220 221 222 223 224
		if (image->segment[i].bufsz > image->segment[i].memsz)
			goto out;
	}

	result = 0;
M
Maneesh Soni 已提交
225 226
out:
	if (result == 0)
227
		*rimage = image;
M
Maneesh Soni 已提交
228
	else
229
		kfree(image);
M
Maneesh Soni 已提交
230

231 232 233 234
	return result;

}

235 236
static void kimage_free_page_list(struct list_head *list);

237
static int kimage_normal_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
238 239
				unsigned long nr_segments,
				struct kexec_segment __user *segments)
240 241 242 243 244 245 246
{
	int result;
	struct kimage *image;

	/* Allocate and initialize a controlling structure */
	image = NULL;
	result = do_kimage_alloc(&image, entry, nr_segments, segments);
M
Maneesh Soni 已提交
247
	if (result)
248
		goto out;
M
Maneesh Soni 已提交
249

250 251 252 253 254 255 256
	/*
	 * 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.
	 */
	result = -ENOMEM;
	image->control_code_page = kimage_alloc_control_pages(image,
257
					   get_order(KEXEC_CONTROL_PAGE_SIZE));
258 259
	if (!image->control_code_page) {
		printk(KERN_ERR "Could not allocate control_code_buffer\n");
260
		goto out_free;
261 262
	}

H
Huang Ying 已提交
263 264 265
	image->swap_page = kimage_alloc_control_pages(image, 0);
	if (!image->swap_page) {
		printk(KERN_ERR "Could not allocate swap buffer\n");
266
		goto out_free;
H
Huang Ying 已提交
267 268
	}

269 270
	*rimage = image;
	return 0;
M
Maneesh Soni 已提交
271

272 273 274 275
out_free:
	kimage_free_page_list(&image->control_pages);
	kfree(image);
out:
276 277 278 279
	return result;
}

static int kimage_crash_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
280
				unsigned long nr_segments,
281
				struct kexec_segment __user *segments)
282 283 284 285 286 287 288 289 290 291 292 293 294 295
{
	int result;
	struct kimage *image;
	unsigned long i;

	image = NULL;
	/* Verify we have a valid entry point */
	if ((entry < crashk_res.start) || (entry > crashk_res.end)) {
		result = -EADDRNOTAVAIL;
		goto out;
	}

	/* Allocate and initialize a controlling structure */
	result = do_kimage_alloc(&image, entry, nr_segments, segments);
M
Maneesh Soni 已提交
296
	if (result)
297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316
		goto out;

	/* Enable the special crash kernel control page
	 * allocation policy.
	 */
	image->control_page = crashk_res.start;
	image->type = KEXEC_TYPE_CRASH;

	/*
	 * 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.
	 */
	result = -EADDRNOTAVAIL;
	for (i = 0; i < nr_segments; i++) {
		unsigned long mstart, mend;
M
Maneesh Soni 已提交
317

318
		mstart = image->segment[i].mem;
319
		mend = mstart + image->segment[i].memsz - 1;
320 321
		/* Ensure we are within the crash kernel limits */
		if ((mstart < crashk_res.start) || (mend > crashk_res.end))
322
			goto out_free;
323 324 325 326 327 328 329 330 331
	}

	/*
	 * Find a location for the control code buffer, and add
	 * the vector of segments so that it's pages will also be
	 * counted as destination pages.
	 */
	result = -ENOMEM;
	image->control_code_page = kimage_alloc_control_pages(image,
332
					   get_order(KEXEC_CONTROL_PAGE_SIZE));
333 334
	if (!image->control_code_page) {
		printk(KERN_ERR "Could not allocate control_code_buffer\n");
335
		goto out_free;
336 337
	}

338 339
	*rimage = image;
	return 0;
M
Maneesh Soni 已提交
340

341 342 343
out_free:
	kfree(image);
out:
344 345 346
	return result;
}

M
Maneesh Soni 已提交
347 348 349
static int kimage_is_destination_range(struct kimage *image,
					unsigned long start,
					unsigned long end)
350 351 352 353 354
{
	unsigned long i;

	for (i = 0; i < image->nr_segments; i++) {
		unsigned long mstart, mend;
M
Maneesh Soni 已提交
355

356
		mstart = image->segment[i].mem;
M
Maneesh Soni 已提交
357 358
		mend = mstart + image->segment[i].memsz;
		if ((end > mstart) && (start < mend))
359 360
			return 1;
	}
M
Maneesh Soni 已提交
361

362 363 364
	return 0;
}

A
Al Viro 已提交
365
static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order)
366 367
{
	struct page *pages;
M
Maneesh Soni 已提交
368

369 370 371 372
	pages = alloc_pages(gfp_mask, order);
	if (pages) {
		unsigned int count, i;
		pages->mapping = NULL;
H
Hugh Dickins 已提交
373
		set_page_private(pages, order);
374
		count = 1 << order;
M
Maneesh Soni 已提交
375
		for (i = 0; i < count; i++)
376 377
			SetPageReserved(pages + i);
	}
M
Maneesh Soni 已提交
378

379 380 381 382 383 384
	return pages;
}

static void kimage_free_pages(struct page *page)
{
	unsigned int order, count, i;
M
Maneesh Soni 已提交
385

H
Hugh Dickins 已提交
386
	order = page_private(page);
387
	count = 1 << order;
M
Maneesh Soni 已提交
388
	for (i = 0; i < count; i++)
389 390 391 392 393 394 395
		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
396

397 398 399 400 401 402 403 404 405
	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
406 407
static struct page *kimage_alloc_normal_control_pages(struct kimage *image,
							unsigned int order)
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
{
	/* 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
434

435 436 437 438 439 440 441 442
		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
443
			      kimage_is_destination_range(image, addr, eaddr)) {
444 445 446
			list_add(&pages->lru, &extra_pages);
			pages = NULL;
		}
M
Maneesh Soni 已提交
447 448
	} while (!pages);

449 450 451 452 453 454 455 456 457 458 459 460 461 462
	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
463
	 * page allocations, and add everything to image->dest_pages.
464 465 466 467 468
	 *
	 * For now it is simpler to just free the pages.
	 */
	kimage_free_page_list(&extra_pages);

M
Maneesh Soni 已提交
469
	return pages;
470 471
}

M
Maneesh Soni 已提交
472 473
static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
						      unsigned int order)
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497
{
	/* 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
498

499 500 501 502
	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
503
	while (hole_end <= crashk_res.end) {
504
		unsigned long i;
M
Maneesh Soni 已提交
505

506
		if (hole_end > KEXEC_CRASH_CONTROL_MEMORY_LIMIT)
507 508
			break;
		/* See if I overlap any of the segments */
M
Maneesh Soni 已提交
509
		for (i = 0; i < image->nr_segments; i++) {
510
			unsigned long mstart, mend;
M
Maneesh Soni 已提交
511

512 513 514 515 516 517 518 519 520 521 522 523 524 525 526
			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
527
	if (pages)
528
		image->control_page = hole_end;
M
Maneesh Soni 已提交
529

530 531 532 533
	return pages;
}


M
Maneesh Soni 已提交
534 535
struct page *kimage_alloc_control_pages(struct kimage *image,
					 unsigned int order)
536 537
{
	struct page *pages = NULL;
M
Maneesh Soni 已提交
538 539

	switch (image->type) {
540 541 542 543 544 545 546
	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
547

548 549 550 551 552
	return pages;
}

static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
{
M
Maneesh Soni 已提交
553
	if (*image->entry != 0)
554
		image->entry++;
M
Maneesh Soni 已提交
555

556 557 558
	if (image->entry == image->last_entry) {
		kimage_entry_t *ind_page;
		struct page *page;
M
Maneesh Soni 已提交
559

560
		page = kimage_alloc_page(image, GFP_KERNEL, KIMAGE_NO_DEST);
M
Maneesh Soni 已提交
561
		if (!page)
562
			return -ENOMEM;
M
Maneesh Soni 已提交
563

564 565 566
		ind_page = page_address(page);
		*image->entry = virt_to_phys(ind_page) | IND_INDIRECTION;
		image->entry = ind_page;
M
Maneesh Soni 已提交
567 568
		image->last_entry = ind_page +
				      ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1);
569 570 571 572
	}
	*image->entry = entry;
	image->entry++;
	*image->entry = 0;
M
Maneesh Soni 已提交
573

574 575 576
	return 0;
}

M
Maneesh Soni 已提交
577 578
static int kimage_set_destination(struct kimage *image,
				   unsigned long destination)
579 580 581 582 583
{
	int result;

	destination &= PAGE_MASK;
	result = kimage_add_entry(image, destination | IND_DESTINATION);
M
Maneesh Soni 已提交
584
	if (result == 0)
585
		image->destination = destination;
M
Maneesh Soni 已提交
586

587 588 589 590 591 592 593 594 595 596
	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
597
	if (result == 0)
598
		image->destination += PAGE_SIZE;
M
Maneesh Soni 已提交
599

600 601 602 603 604 605 606 607 608
	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
609
	/* Walk through and free any unusable pages I have cached */
610 611 612
	kimage_free_page_list(&image->unuseable_pages);

}
613
static void kimage_terminate(struct kimage *image)
614
{
M
Maneesh Soni 已提交
615
	if (*image->entry != 0)
616
		image->entry++;
M
Maneesh Soni 已提交
617

618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
	*image->entry = IND_DONE;
}

#define for_each_kimage_entry(image, ptr, entry) \
	for (ptr = &image->head; (entry = *ptr) && !(entry & IND_DONE); \
		ptr = (entry & IND_INDIRECTION)? \
			phys_to_virt((entry & PAGE_MASK)): ptr +1)

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
641

642 643 644 645
	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
646
			if (ind & IND_INDIRECTION)
647 648 649 650 651 652
				kimage_free_entry(ind);
			/* Save this indirection page until we are
			 * done with it.
			 */
			ind = entry;
		}
M
Maneesh Soni 已提交
653
		else if (entry & IND_SOURCE)
654 655 656
			kimage_free_entry(entry);
	}
	/* Free the final indirection page */
M
Maneesh Soni 已提交
657
	if (ind & IND_INDIRECTION)
658 659 660 661 662 663 664 665 666 667
		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);
	kfree(image);
}

M
Maneesh Soni 已提交
668 669
static kimage_entry_t *kimage_dst_used(struct kimage *image,
					unsigned long page)
670 671 672 673 674
{
	kimage_entry_t *ptr, entry;
	unsigned long destination = 0;

	for_each_kimage_entry(image, ptr, entry) {
M
Maneesh Soni 已提交
675
		if (entry & IND_DESTINATION)
676 677
			destination = entry & PAGE_MASK;
		else if (entry & IND_SOURCE) {
M
Maneesh Soni 已提交
678
			if (page == destination)
679 680 681 682
				return ptr;
			destination += PAGE_SIZE;
		}
	}
M
Maneesh Soni 已提交
683

684
	return NULL;
685 686
}

M
Maneesh Soni 已提交
687
static struct page *kimage_alloc_page(struct kimage *image,
A
Al Viro 已提交
688
					gfp_t gfp_mask,
M
Maneesh Soni 已提交
689
					unsigned long destination)
690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728
{
	/*
	 * 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
729
		if (!page)
730
			return NULL;
731
		/* If the page cannot be used file it away */
M
Maneesh Soni 已提交
732 733
		if (page_to_pfn(page) >
				(KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) {
734 735 736 737 738 739 740 741 742 743
			list_add(&page->lru, &image->unuseable_pages);
			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
744 745
		if (!kimage_is_destination_range(image, addr,
						  addr + PAGE_SIZE))
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764
			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
765 766
			 * destination page, so return it if it's
			 * gfp_flags honor the ones passed in.
767
			 */
768 769 770 771 772
			if (!(gfp_mask & __GFP_HIGHMEM) &&
			    PageHighMem(old_page)) {
				kimage_free_pages(old_page);
				continue;
			}
773 774 775 776 777 778 779 780 781 782 783
			addr = old_addr;
			page = old_page;
			break;
		}
		else {
			/* 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
784

785 786 787 788
	return page;
}

static int kimage_load_normal_segment(struct kimage *image,
M
Maneesh Soni 已提交
789
					 struct kexec_segment *segment)
790 791
{
	unsigned long maddr;
792
	size_t ubytes, mbytes;
793
	int result;
794
	unsigned char __user *buf;
795 796 797 798 799 800 801 802

	result = 0;
	buf = segment->buf;
	ubytes = segment->bufsz;
	mbytes = segment->memsz;
	maddr = segment->mem;

	result = kimage_set_destination(image, maddr);
M
Maneesh Soni 已提交
803
	if (result < 0)
804
		goto out;
M
Maneesh Soni 已提交
805 806

	while (mbytes) {
807 808 809
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
810

811
		page = kimage_alloc_page(image, GFP_HIGHUSER, maddr);
812
		if (!page) {
813 814 815
			result  = -ENOMEM;
			goto out;
		}
M
Maneesh Soni 已提交
816 817 818
		result = kimage_add_page(image, page_to_pfn(page)
								<< PAGE_SHIFT);
		if (result < 0)
819
			goto out;
M
Maneesh Soni 已提交
820

821 822
		ptr = kmap(page);
		/* Start with a clear page */
823
		clear_page(ptr);
824
		ptr += maddr & ~PAGE_MASK;
825 826 827
		mchunk = min_t(size_t, mbytes,
				PAGE_SIZE - (maddr & ~PAGE_MASK));
		uchunk = min(ubytes, mchunk);
M
Maneesh Soni 已提交
828

829 830 831
		result = copy_from_user(ptr, buf, uchunk);
		kunmap(page);
		if (result) {
832
			result = -EFAULT;
833 834 835 836 837 838 839
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
		buf    += mchunk;
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
840
out:
841 842 843 844
	return result;
}

static int kimage_load_crash_segment(struct kimage *image,
M
Maneesh Soni 已提交
845
					struct kexec_segment *segment)
846 847 848 849 850 851
{
	/* 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;
852
	size_t ubytes, mbytes;
853
	int result;
854
	unsigned char __user *buf;
855 856 857 858 859 860

	result = 0;
	buf = segment->buf;
	ubytes = segment->bufsz;
	mbytes = segment->memsz;
	maddr = segment->mem;
M
Maneesh Soni 已提交
861
	while (mbytes) {
862 863 864
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
865

866
		page = pfn_to_page(maddr >> PAGE_SHIFT);
867
		if (!page) {
868 869 870 871 872
			result  = -ENOMEM;
			goto out;
		}
		ptr = kmap(page);
		ptr += maddr & ~PAGE_MASK;
873 874 875 876
		mchunk = min_t(size_t, mbytes,
				PAGE_SIZE - (maddr & ~PAGE_MASK));
		uchunk = min(ubytes, mchunk);
		if (mchunk > uchunk) {
877 878 879 880
			/* 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
881
		kexec_flush_icache_page(page);
882 883
		kunmap(page);
		if (result) {
884
			result = -EFAULT;
885 886 887 888 889 890 891
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
		buf    += mchunk;
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
892
out:
893 894 895 896
	return result;
}

static int kimage_load_segment(struct kimage *image,
M
Maneesh Soni 已提交
897
				struct kexec_segment *segment)
898 899
{
	int result = -ENOMEM;
M
Maneesh Soni 已提交
900 901

	switch (image->type) {
902 903 904 905 906 907 908
	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
909

910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926
	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
927
 *   and then copies the image to it's final destination.  And
928 929 930 931 932
 *   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.
 */
933 934
struct kimage *kexec_image;
struct kimage *kexec_crash_image;
935
int kexec_load_disabled;
936 937

static DEFINE_MUTEX(kexec_mutex);
938

939 940
SYSCALL_DEFINE4(kexec_load, unsigned long, entry, unsigned long, nr_segments,
		struct kexec_segment __user *, segments, unsigned long, flags)
941 942 943 944 945
{
	struct kimage **dest_image, *image;
	int result;

	/* We only trust the superuser with rebooting the system. */
946
	if (!capable(CAP_SYS_BOOT) || kexec_load_disabled)
947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977
		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.
	 */
978
	if (!mutex_trylock(&kexec_mutex))
979
		return -EBUSY;
M
Maneesh Soni 已提交
980

981
	dest_image = &kexec_image;
M
Maneesh Soni 已提交
982
	if (flags & KEXEC_ON_CRASH)
983 984 985
		dest_image = &kexec_crash_image;
	if (nr_segments > 0) {
		unsigned long i;
M
Maneesh Soni 已提交
986

987
		/* Loading another kernel to reboot into */
M
Maneesh Soni 已提交
988 989 990
		if ((flags & KEXEC_ON_CRASH) == 0)
			result = kimage_normal_alloc(&image, entry,
							nr_segments, segments);
991 992 993 994 995 996
		/* 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));
M
Maneesh Soni 已提交
997 998
			result = kimage_crash_alloc(&image, entry,
						     nr_segments, segments);
999
			crash_map_reserved_pages();
1000
		}
M
Maneesh Soni 已提交
1001
		if (result)
1002
			goto out;
M
Maneesh Soni 已提交
1003

H
Huang Ying 已提交
1004 1005
		if (flags & KEXEC_PRESERVE_CONTEXT)
			image->preserve_context = 1;
1006
		result = machine_kexec_prepare(image);
M
Maneesh Soni 已提交
1007
		if (result)
1008
			goto out;
M
Maneesh Soni 已提交
1009 1010

		for (i = 0; i < nr_segments; i++) {
1011
			result = kimage_load_segment(image, &image->segment[i]);
M
Maneesh Soni 已提交
1012
			if (result)
1013 1014
				goto out;
		}
1015
		kimage_terminate(image);
1016 1017
		if (flags & KEXEC_ON_CRASH)
			crash_unmap_reserved_pages();
1018 1019 1020 1021
	}
	/* Install the new kernel, and  Uninstall the old */
	image = xchg(dest_image, image);

M
Maneesh Soni 已提交
1022
out:
1023
	mutex_unlock(&kexec_mutex);
1024
	kimage_free(image);
M
Maneesh Soni 已提交
1025

1026 1027 1028
	return result;
}

1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
/*
 * 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)
{}

1041 1042
#ifdef CONFIG_COMPAT
asmlinkage long compat_sys_kexec_load(unsigned long entry,
M
Maneesh Soni 已提交
1043 1044 1045
				unsigned long nr_segments,
				struct compat_kexec_segment __user *segments,
				unsigned long flags)
1046 1047 1048 1049 1050 1051 1052 1053
{
	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
1054
	if ((flags & KEXEC_ARCH_MASK) == KEXEC_ARCH_DEFAULT)
1055 1056
		return -EINVAL;

M
Maneesh Soni 已提交
1057
	if (nr_segments > KEXEC_SEGMENT_MAX)
1058 1059 1060 1061 1062
		return -EINVAL;

	ksegments = compat_alloc_user_space(nr_segments * sizeof(out));
	for (i=0; i < nr_segments; i++) {
		result = copy_from_user(&in, &segments[i], sizeof(in));
M
Maneesh Soni 已提交
1063
		if (result)
1064 1065 1066 1067 1068 1069 1070 1071
			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
1072
		if (result)
1073 1074 1075 1076 1077 1078 1079
			return -EFAULT;
	}

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

1080
void crash_kexec(struct pt_regs *regs)
1081
{
1082
	/* Take the kexec_mutex here to prevent sys_kexec_load
1083 1084 1085 1086 1087 1088 1089
	 * 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...
	 */
1090
	if (mutex_trylock(&kexec_mutex)) {
1091
		if (kexec_crash_image) {
1092
			struct pt_regs fixed_regs;
1093

1094
			crash_setup_regs(&fixed_regs, regs);
K
Ken'ichi Ohmichi 已提交
1095
			crash_save_vmcoreinfo();
1096
			machine_crash_shutdown(&fixed_regs);
1097
			machine_kexec(kexec_crash_image);
1098
		}
1099
		mutex_unlock(&kexec_mutex);
1100 1101
	}
}
1102

1103 1104
size_t crash_get_memory_size(void)
{
1105
	size_t size = 0;
1106
	mutex_lock(&kexec_mutex);
1107
	if (crashk_res.end != crashk_res.start)
1108
		size = resource_size(&crashk_res);
1109 1110 1111 1112
	mutex_unlock(&kexec_mutex);
	return size;
}

1113 1114
void __weak crash_free_reserved_phys_range(unsigned long begin,
					   unsigned long end)
1115 1116 1117
{
	unsigned long addr;

1118 1119
	for (addr = begin; addr < end; addr += PAGE_SIZE)
		free_reserved_page(pfn_to_page(addr >> PAGE_SHIFT));
1120 1121 1122 1123 1124 1125
}

int crash_shrink_memory(unsigned long new_size)
{
	int ret = 0;
	unsigned long start, end;
1126
	unsigned long old_size;
1127
	struct resource *ram_res;
1128 1129 1130 1131 1132 1133 1134 1135 1136

	mutex_lock(&kexec_mutex);

	if (kexec_crash_image) {
		ret = -ENOENT;
		goto unlock;
	}
	start = crashk_res.start;
	end = crashk_res.end;
1137 1138 1139
	old_size = (end == 0) ? 0 : end - start + 1;
	if (new_size >= old_size) {
		ret = (new_size == old_size) ? 0 : -EINVAL;
1140 1141 1142
		goto unlock;
	}

1143 1144 1145 1146 1147 1148
	ram_res = kzalloc(sizeof(*ram_res), GFP_KERNEL);
	if (!ram_res) {
		ret = -ENOMEM;
		goto unlock;
	}

1149 1150
	start = roundup(start, KEXEC_CRASH_MEM_ALIGN);
	end = roundup(start + new_size, KEXEC_CRASH_MEM_ALIGN);
1151

1152
	crash_map_reserved_pages();
1153
	crash_free_reserved_phys_range(end, crashk_res.end);
1154

1155
	if ((start == end) && (crashk_res.parent != NULL))
1156
		release_resource(&crashk_res);
1157 1158 1159 1160 1161 1162

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

1163
	crashk_res.end = end - 1;
1164 1165

	insert_resource(&iomem_resource, ram_res);
1166
	crash_unmap_reserved_pages();
1167 1168 1169 1170 1171 1172

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

1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205
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;

1206
	if ((cpu < 0) || (cpu >= nr_cpu_ids))
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220
		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.
	 */
	buf = (u32*)per_cpu_ptr(crash_notes, cpu);
	if (!buf)
		return;
	memset(&prstatus, 0, sizeof(prstatus));
	prstatus.pr_pid = current->pid;
T
Tejun Heo 已提交
1221
	elf_core_copy_kernel_regs(&prstatus.pr_reg, regs);
1222 1223
	buf = append_elf_note(buf, KEXEC_CORE_NOTE_NAME, NT_PRSTATUS,
		      	      &prstatus, sizeof(prstatus));
1224 1225 1226
	final_note(buf);
}

1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238
static int __init crash_notes_memory_init(void)
{
	/* Allocate memory for saving cpu registers. */
	crash_notes = alloc_percpu(note_buf_t);
	if (!crash_notes) {
		printk("Kexec: Memory allocation for saving cpu register"
		" states failed\n");
		return -ENOMEM;
	}
	return 0;
}
module_init(crash_notes_memory_init)
K
Ken'ichi Ohmichi 已提交
1239

B
Bernhard Walle 已提交
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311

/*
 * 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.
 */
static int __init parse_crashkernel_mem(char 			*cmdline,
					unsigned long long	system_ram,
					unsigned long long	*crash_size,
					unsigned long long	*crash_base)
{
	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) {
			pr_warning("crashkernel: Memory value expected\n");
			return -EINVAL;
		}
		cur = tmp;
		if (*cur != '-') {
			pr_warning("crashkernel: '-' expected\n");
			return -EINVAL;
		}
		cur++;

		/* if no ':' is here, than we read the end */
		if (*cur != ':') {
			end = memparse(cur, &tmp);
			if (cur == tmp) {
				pr_warning("crashkernel: Memory "
						"value expected\n");
				return -EINVAL;
			}
			cur = tmp;
			if (end <= start) {
				pr_warning("crashkernel: end <= start\n");
				return -EINVAL;
			}
		}

		if (*cur != ':') {
			pr_warning("crashkernel: ':' expected\n");
			return -EINVAL;
		}
		cur++;

		size = memparse(cur, &tmp);
		if (cur == tmp) {
			pr_warning("Memory value expected\n");
			return -EINVAL;
		}
		cur = tmp;
		if (size >= system_ram) {
			pr_warning("crashkernel: invalid size\n");
			return -EINVAL;
		}

		/* match ? */
1312
		if (system_ram >= start && system_ram < end) {
B
Bernhard Walle 已提交
1313 1314 1315 1316 1317 1318
			*crash_size = size;
			break;
		}
	} while (*cur++ == ',');

	if (*crash_size > 0) {
1319
		while (*cur && *cur != ' ' && *cur != '@')
B
Bernhard Walle 已提交
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
			cur++;
		if (*cur == '@') {
			cur++;
			*crash_base = memparse(cur, &tmp);
			if (cur == tmp) {
				pr_warning("Memory value expected "
						"after '@'\n");
				return -EINVAL;
			}
		}
	}

	return 0;
}

/*
 * That function parses "simple" (old) crashkernel command lines like
 *
 * 	crashkernel=size[@offset]
 *
 * It returns 0 on success and -EINVAL on failure.
 */
static int __init parse_crashkernel_simple(char 		*cmdline,
					   unsigned long long 	*crash_size,
					   unsigned long long 	*crash_base)
{
	char *cur = cmdline;

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

	if (*cur == '@')
		*crash_base = memparse(cur+1, &cur);
1356 1357 1358 1359
	else if (*cur != ' ' && *cur != '\0') {
		pr_warning("crashkernel: unrecognized char\n");
		return -EINVAL;
	}
B
Bernhard Walle 已提交
1360 1361 1362 1363

	return 0;
}

1364 1365 1366 1367 1368 1369 1370 1371 1372
#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
1373
/*
1374 1375 1376 1377 1378
 * 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
1379
 */
1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447
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;
}

1448
static int __init __parse_crashkernel(char *cmdline,
B
Bernhard Walle 已提交
1449 1450
			     unsigned long long system_ram,
			     unsigned long long *crash_size,
1451
			     unsigned long long *crash_base,
1452 1453
			     const char *name,
			     const char *suffix)
B
Bernhard Walle 已提交
1454 1455
{
	char	*first_colon, *first_space;
1456
	char	*ck_cmdline;
B
Bernhard Walle 已提交
1457 1458 1459 1460 1461

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

1462
	ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
B
Bernhard Walle 已提交
1463 1464 1465 1466

	if (!ck_cmdline)
		return -EINVAL;

1467
	ck_cmdline += strlen(name);
B
Bernhard Walle 已提交
1468

1469 1470 1471
	if (suffix)
		return parse_crashkernel_suffix(ck_cmdline, crash_size,
				crash_base, suffix);
B
Bernhard Walle 已提交
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481
	/*
	 * 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
1482
	return parse_crashkernel_simple(ck_cmdline, crash_size, crash_base);
B
Bernhard Walle 已提交
1483 1484
}

1485 1486 1487 1488
/*
 * That function is the entry point for command line parsing and should be
 * called from the arch-specific code.
 */
1489 1490 1491 1492 1493 1494
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,
1495
					"crashkernel=", NULL);
1496
}
1497 1498 1499 1500 1501 1502 1503

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,
1504
				"crashkernel=", suffix_tbl[SUFFIX_HIGH]);
1505
}
1506 1507 1508 1509 1510 1511 1512

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,
1513
				"crashkernel=", suffix_tbl[SUFFIX_LOW]);
1514
}
B
Bernhard Walle 已提交
1515

1516
static void update_vmcoreinfo_note(void)
K
Ken'ichi Ohmichi 已提交
1517
{
1518
	u32 *buf = vmcoreinfo_note;
K
Ken'ichi Ohmichi 已提交
1519 1520 1521 1522 1523 1524 1525 1526

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

1527 1528
void crash_save_vmcoreinfo(void)
{
1529
	vmcoreinfo_append_str("CRASHTIME=%ld\n", get_seconds());
1530 1531 1532
	update_vmcoreinfo_note();
}

K
Ken'ichi Ohmichi 已提交
1533 1534 1535 1536
void vmcoreinfo_append_str(const char *fmt, ...)
{
	va_list args;
	char buf[0x50];
1537
	size_t r;
K
Ken'ichi Ohmichi 已提交
1538 1539

	va_start(args, fmt);
1540
	r = vscnprintf(buf, sizeof(buf), fmt, args);
K
Ken'ichi Ohmichi 已提交
1541 1542
	va_end(args);

1543
	r = min(r, vmcoreinfo_max_size - vmcoreinfo_size);
K
Ken'ichi Ohmichi 已提交
1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563

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

	vmcoreinfo_size += r;
}

/*
 * provide an empty default implementation here -- architecture
 * code may override this
 */
void __attribute__ ((weak)) arch_crash_save_vmcoreinfo(void)
{}

unsigned long __attribute__ ((weak)) paddr_vmcoreinfo_note(void)
{
	return __pa((unsigned long)(char *)&vmcoreinfo_note);
}

static int __init crash_save_vmcoreinfo_init(void)
{
1564 1565
	VMCOREINFO_OSRELEASE(init_uts_ns.name.release);
	VMCOREINFO_PAGESIZE(PAGE_SIZE);
K
Ken'ichi Ohmichi 已提交
1566

1567 1568
	VMCOREINFO_SYMBOL(init_uts_ns);
	VMCOREINFO_SYMBOL(node_online_map);
1569
#ifdef CONFIG_MMU
1570
	VMCOREINFO_SYMBOL(swapper_pg_dir);
1571
#endif
1572
	VMCOREINFO_SYMBOL(_stext);
1573
	VMCOREINFO_SYMBOL(vmap_area_list);
K
Ken'ichi Ohmichi 已提交
1574 1575

#ifndef CONFIG_NEED_MULTIPLE_NODES
1576 1577
	VMCOREINFO_SYMBOL(mem_map);
	VMCOREINFO_SYMBOL(contig_page_data);
K
Ken'ichi Ohmichi 已提交
1578 1579
#endif
#ifdef CONFIG_SPARSEMEM
1580 1581
	VMCOREINFO_SYMBOL(mem_section);
	VMCOREINFO_LENGTH(mem_section, NR_SECTION_ROOTS);
1582
	VMCOREINFO_STRUCT_SIZE(mem_section);
1583
	VMCOREINFO_OFFSET(mem_section, section_mem_map);
K
Ken'ichi Ohmichi 已提交
1584
#endif
1585 1586 1587 1588 1589 1590
	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);
1591 1592 1593 1594
	VMCOREINFO_OFFSET(page, flags);
	VMCOREINFO_OFFSET(page, _count);
	VMCOREINFO_OFFSET(page, mapping);
	VMCOREINFO_OFFSET(page, lru);
1595 1596
	VMCOREINFO_OFFSET(page, _mapcount);
	VMCOREINFO_OFFSET(page, private);
1597 1598
	VMCOREINFO_OFFSET(pglist_data, node_zones);
	VMCOREINFO_OFFSET(pglist_data, nr_zones);
K
Ken'ichi Ohmichi 已提交
1599
#ifdef CONFIG_FLAT_NODE_MEM_MAP
1600
	VMCOREINFO_OFFSET(pglist_data, node_mem_map);
K
Ken'ichi Ohmichi 已提交
1601
#endif
1602 1603 1604 1605 1606 1607 1608 1609 1610
	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);
1611 1612
	VMCOREINFO_OFFSET(vmap_area, va_start);
	VMCOREINFO_OFFSET(vmap_area, list);
1613
	VMCOREINFO_LENGTH(zone.free_area, MAX_ORDER);
1614
	log_buf_kexec_setup();
1615
	VMCOREINFO_LENGTH(free_area.free_list, MIGRATE_TYPES);
1616
	VMCOREINFO_NUMBER(NR_FREE_PAGES);
1617 1618 1619
	VMCOREINFO_NUMBER(PG_lru);
	VMCOREINFO_NUMBER(PG_private);
	VMCOREINFO_NUMBER(PG_swapcache);
1620
	VMCOREINFO_NUMBER(PG_slab);
1621 1622 1623
#ifdef CONFIG_MEMORY_FAILURE
	VMCOREINFO_NUMBER(PG_hwpoison);
#endif
1624
	VMCOREINFO_NUMBER(PAGE_BUDDY_MAPCOUNT_VALUE);
K
Ken'ichi Ohmichi 已提交
1625 1626

	arch_crash_save_vmcoreinfo();
1627
	update_vmcoreinfo_note();
K
Ken'ichi Ohmichi 已提交
1628 1629 1630 1631 1632

	return 0;
}

module_init(crash_save_vmcoreinfo_init)
H
Huang Ying 已提交
1633

1634 1635 1636
/*
 * Move into place and start executing a preloaded standalone
 * executable.  If nothing was preloaded return an error.
H
Huang Ying 已提交
1637 1638 1639 1640 1641
 */
int kernel_kexec(void)
{
	int error = 0;

1642
	if (!mutex_trylock(&kexec_mutex))
H
Huang Ying 已提交
1643 1644 1645 1646 1647 1648 1649
		return -EBUSY;
	if (!kexec_image) {
		error = -EINVAL;
		goto Unlock;
	}

#ifdef CONFIG_KEXEC_JUMP
1650
	if (kexec_image->preserve_context) {
1651
		lock_system_sleep();
1652 1653 1654 1655 1656 1657 1658
		pm_prepare_console();
		error = freeze_processes();
		if (error) {
			error = -EBUSY;
			goto Restore_console;
		}
		suspend_console();
1659
		error = dpm_suspend_start(PMSG_FREEZE);
1660 1661
		if (error)
			goto Resume_console;
1662
		/* At this point, dpm_suspend_start() has been called,
1663 1664
		 * but *not* dpm_suspend_end(). We *must* call
		 * dpm_suspend_end() now.  Otherwise, drivers for
1665 1666 1667 1668
		 * some devices (e.g. interrupt controllers) become
		 * desynchronized with the actual state of the
		 * hardware at resume time, and evil weirdness ensues.
		 */
1669
		error = dpm_suspend_end(PMSG_FREEZE);
1670
		if (error)
1671 1672 1673 1674
			goto Resume_devices;
		error = disable_nonboot_cpus();
		if (error)
			goto Enable_cpus;
1675
		local_irq_disable();
1676
		error = syscore_suspend();
1677
		if (error)
1678
			goto Enable_irqs;
1679
	} else
H
Huang Ying 已提交
1680
#endif
1681
	{
1682
		kexec_in_progress = true;
1683
		kernel_restart_prepare(NULL);
V
Vivek Goyal 已提交
1684
		migrate_to_reboot_cpu();
H
Huang Ying 已提交
1685 1686 1687 1688 1689 1690 1691
		printk(KERN_EMERG "Starting new kernel\n");
		machine_shutdown();
	}

	machine_kexec(kexec_image);

#ifdef CONFIG_KEXEC_JUMP
1692
	if (kexec_image->preserve_context) {
1693
		syscore_resume();
1694
 Enable_irqs:
H
Huang Ying 已提交
1695
		local_irq_enable();
1696
 Enable_cpus:
1697
		enable_nonboot_cpus();
1698
		dpm_resume_start(PMSG_RESTORE);
1699
 Resume_devices:
1700
		dpm_resume_end(PMSG_RESTORE);
1701 1702 1703 1704 1705
 Resume_console:
		resume_console();
		thaw_processes();
 Restore_console:
		pm_restore_console();
1706
		unlock_system_sleep();
H
Huang Ying 已提交
1707
	}
1708
#endif
H
Huang Ying 已提交
1709 1710

 Unlock:
1711
	mutex_unlock(&kexec_mutex);
H
Huang Ying 已提交
1712 1713
	return error;
}