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

64 65
int kexec_should_crash(struct task_struct *p)
{
66
	if (in_interrupt() || !p->pid || is_global_init(p) || panic_on_oops)
67 68 69 70
		return 1;
	return 0;
}

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
/*
 * 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
87
 * is given by KEXEC_CONTROL_PAGE_SIZE.  In the best case only a single
88 89 90 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
 * 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
117 118 119
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
120
				       gfp_t gfp_mask,
M
Maneesh Soni 已提交
121
				       unsigned long dest);
122 123

static int do_kimage_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
124 125
	                    unsigned long nr_segments,
                            struct kexec_segment __user *segments)
126 127 128 129 130 131 132 133
{
	size_t segment_bytes;
	struct kimage *image;
	unsigned long i;
	int result;

	/* Allocate a controlling structure */
	result = -ENOMEM;
134
	image = kzalloc(sizeof(*image), GFP_KERNEL);
M
Maneesh Soni 已提交
135
	if (!image)
136
		goto out;
M
Maneesh Soni 已提交
137

138 139 140 141 142 143 144 145 146 147 148 149 150
	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
151
	/* Initialize the list of unusable pages */
152 153 154 155 156 157
	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);
158 159
	if (result) {
		result = -EFAULT;
160
		goto out;
161
	}
162 163 164 165 166 167 168 169

	/*
	 * 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
170
	 * the destination addresses are page aligned.  Too many
171 172 173 174 175 176 177 178
	 * 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
179

180 181 182 183 184 185 186 187 188 189 190 191 192 193
		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
194
	for (i = 0; i < nr_segments; i++) {
195 196
		unsigned long mstart, mend;
		unsigned long j;
M
Maneesh Soni 已提交
197

198 199
		mstart = image->segment[i].mem;
		mend   = mstart + image->segment[i].memsz;
M
Maneesh Soni 已提交
200
		for (j = 0; j < i; j++) {
201 202 203 204 205 206 207 208 209 210 211 212 213 214 215
			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
216
	for (i = 0; i < nr_segments; i++) {
217 218 219 220 221
		if (image->segment[i].bufsz > image->segment[i].memsz)
			goto out;
	}

	result = 0;
M
Maneesh Soni 已提交
222 223
out:
	if (result == 0)
224
		*rimage = image;
M
Maneesh Soni 已提交
225
	else
226
		kfree(image);
M
Maneesh Soni 已提交
227

228 229 230 231
	return result;

}

232 233
static void kimage_free_page_list(struct list_head *list);

234
static int kimage_normal_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
235 236
				unsigned long nr_segments,
				struct kexec_segment __user *segments)
237 238 239 240 241 242 243
{
	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
244
	if (result)
245
		goto out;
M
Maneesh Soni 已提交
246

247 248 249 250 251 252 253
	/*
	 * 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,
254
					   get_order(KEXEC_CONTROL_PAGE_SIZE));
255 256
	if (!image->control_code_page) {
		printk(KERN_ERR "Could not allocate control_code_buffer\n");
257
		goto out_free;
258 259
	}

H
Huang Ying 已提交
260 261 262
	image->swap_page = kimage_alloc_control_pages(image, 0);
	if (!image->swap_page) {
		printk(KERN_ERR "Could not allocate swap buffer\n");
263
		goto out_free;
H
Huang Ying 已提交
264 265
	}

266 267
	*rimage = image;
	return 0;
M
Maneesh Soni 已提交
268

269 270 271 272
out_free:
	kimage_free_page_list(&image->control_pages);
	kfree(image);
out:
273 274 275 276
	return result;
}

static int kimage_crash_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
277
				unsigned long nr_segments,
278
				struct kexec_segment __user *segments)
279 280 281 282 283 284 285 286 287 288 289 290 291 292
{
	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
293
	if (result)
294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313
		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
314

315
		mstart = image->segment[i].mem;
316
		mend = mstart + image->segment[i].memsz - 1;
317 318
		/* Ensure we are within the crash kernel limits */
		if ((mstart < crashk_res.start) || (mend > crashk_res.end))
319
			goto out_free;
320 321 322 323 324 325 326 327 328
	}

	/*
	 * 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,
329
					   get_order(KEXEC_CONTROL_PAGE_SIZE));
330 331
	if (!image->control_code_page) {
		printk(KERN_ERR "Could not allocate control_code_buffer\n");
332
		goto out_free;
333 334
	}

335 336
	*rimage = image;
	return 0;
M
Maneesh Soni 已提交
337

338 339 340
out_free:
	kfree(image);
out:
341 342 343
	return result;
}

M
Maneesh Soni 已提交
344 345 346
static int kimage_is_destination_range(struct kimage *image,
					unsigned long start,
					unsigned long end)
347 348 349 350 351
{
	unsigned long i;

	for (i = 0; i < image->nr_segments; i++) {
		unsigned long mstart, mend;
M
Maneesh Soni 已提交
352

353
		mstart = image->segment[i].mem;
M
Maneesh Soni 已提交
354 355
		mend = mstart + image->segment[i].memsz;
		if ((end > mstart) && (start < mend))
356 357
			return 1;
	}
M
Maneesh Soni 已提交
358

359 360 361
	return 0;
}

A
Al Viro 已提交
362
static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order)
363 364
{
	struct page *pages;
M
Maneesh Soni 已提交
365

366 367 368 369
	pages = alloc_pages(gfp_mask, order);
	if (pages) {
		unsigned int count, i;
		pages->mapping = NULL;
H
Hugh Dickins 已提交
370
		set_page_private(pages, order);
371
		count = 1 << order;
M
Maneesh Soni 已提交
372
		for (i = 0; i < count; i++)
373 374
			SetPageReserved(pages + i);
	}
M
Maneesh Soni 已提交
375

376 377 378 379 380 381
	return pages;
}

static void kimage_free_pages(struct page *page)
{
	unsigned int order, count, i;
M
Maneesh Soni 已提交
382

H
Hugh Dickins 已提交
383
	order = page_private(page);
384
	count = 1 << order;
M
Maneesh Soni 已提交
385
	for (i = 0; i < count; i++)
386 387 388 389 390 391 392
		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
393

394 395 396 397 398 399 400 401 402
	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
403 404
static struct page *kimage_alloc_normal_control_pages(struct kimage *image,
							unsigned int order)
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
{
	/* 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
431

432 433 434 435 436 437 438 439
		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
440
			      kimage_is_destination_range(image, addr, eaddr)) {
441 442 443
			list_add(&pages->lru, &extra_pages);
			pages = NULL;
		}
M
Maneesh Soni 已提交
444 445
	} while (!pages);

446 447 448 449 450 451 452 453 454 455 456 457 458 459
	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
460
	 * page allocations, and add everything to image->dest_pages.
461 462 463 464 465
	 *
	 * For now it is simpler to just free the pages.
	 */
	kimage_free_page_list(&extra_pages);

M
Maneesh Soni 已提交
466
	return pages;
467 468
}

M
Maneesh Soni 已提交
469 470
static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
						      unsigned int order)
471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494
{
	/* 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
495

496 497 498 499
	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
500
	while (hole_end <= crashk_res.end) {
501
		unsigned long i;
M
Maneesh Soni 已提交
502

503
		if (hole_end > KEXEC_CRASH_CONTROL_MEMORY_LIMIT)
504 505
			break;
		/* See if I overlap any of the segments */
M
Maneesh Soni 已提交
506
		for (i = 0; i < image->nr_segments; i++) {
507
			unsigned long mstart, mend;
M
Maneesh Soni 已提交
508

509 510 511 512 513 514 515 516 517 518 519 520 521 522 523
			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
524
	if (pages)
525
		image->control_page = hole_end;
M
Maneesh Soni 已提交
526

527 528 529 530
	return pages;
}


M
Maneesh Soni 已提交
531 532
struct page *kimage_alloc_control_pages(struct kimage *image,
					 unsigned int order)
533 534
{
	struct page *pages = NULL;
M
Maneesh Soni 已提交
535 536

	switch (image->type) {
537 538 539 540 541 542 543
	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
544

545 546 547 548 549
	return pages;
}

static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
{
M
Maneesh Soni 已提交
550
	if (*image->entry != 0)
551
		image->entry++;
M
Maneesh Soni 已提交
552

553 554 555
	if (image->entry == image->last_entry) {
		kimage_entry_t *ind_page;
		struct page *page;
M
Maneesh Soni 已提交
556

557
		page = kimage_alloc_page(image, GFP_KERNEL, KIMAGE_NO_DEST);
M
Maneesh Soni 已提交
558
		if (!page)
559
			return -ENOMEM;
M
Maneesh Soni 已提交
560

561 562 563
		ind_page = page_address(page);
		*image->entry = virt_to_phys(ind_page) | IND_INDIRECTION;
		image->entry = ind_page;
M
Maneesh Soni 已提交
564 565
		image->last_entry = ind_page +
				      ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1);
566 567 568 569
	}
	*image->entry = entry;
	image->entry++;
	*image->entry = 0;
M
Maneesh Soni 已提交
570

571 572 573
	return 0;
}

M
Maneesh Soni 已提交
574 575
static int kimage_set_destination(struct kimage *image,
				   unsigned long destination)
576 577 578 579 580
{
	int result;

	destination &= PAGE_MASK;
	result = kimage_add_entry(image, destination | IND_DESTINATION);
M
Maneesh Soni 已提交
581
	if (result == 0)
582
		image->destination = destination;
M
Maneesh Soni 已提交
583

584 585 586 587 588 589 590 591 592 593
	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
594
	if (result == 0)
595
		image->destination += PAGE_SIZE;
M
Maneesh Soni 已提交
596

597 598 599 600 601 602 603 604 605
	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
606
	/* Walk through and free any unusable pages I have cached */
607 608 609
	kimage_free_page_list(&image->unuseable_pages);

}
610
static void kimage_terminate(struct kimage *image)
611
{
M
Maneesh Soni 已提交
612
	if (*image->entry != 0)
613
		image->entry++;
M
Maneesh Soni 已提交
614

615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637
	*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
638

639 640 641 642
	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
643
			if (ind & IND_INDIRECTION)
644 645 646 647 648 649
				kimage_free_entry(ind);
			/* Save this indirection page until we are
			 * done with it.
			 */
			ind = entry;
		}
M
Maneesh Soni 已提交
650
		else if (entry & IND_SOURCE)
651 652 653
			kimage_free_entry(entry);
	}
	/* Free the final indirection page */
M
Maneesh Soni 已提交
654
	if (ind & IND_INDIRECTION)
655 656 657 658 659 660 661 662 663 664
		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
665 666
static kimage_entry_t *kimage_dst_used(struct kimage *image,
					unsigned long page)
667 668 669 670 671
{
	kimage_entry_t *ptr, entry;
	unsigned long destination = 0;

	for_each_kimage_entry(image, ptr, entry) {
M
Maneesh Soni 已提交
672
		if (entry & IND_DESTINATION)
673 674
			destination = entry & PAGE_MASK;
		else if (entry & IND_SOURCE) {
M
Maneesh Soni 已提交
675
			if (page == destination)
676 677 678 679
				return ptr;
			destination += PAGE_SIZE;
		}
	}
M
Maneesh Soni 已提交
680

681
	return NULL;
682 683
}

M
Maneesh Soni 已提交
684
static struct page *kimage_alloc_page(struct kimage *image,
A
Al Viro 已提交
685
					gfp_t gfp_mask,
M
Maneesh Soni 已提交
686
					unsigned long destination)
687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
{
	/*
	 * 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
726
		if (!page)
727
			return NULL;
728
		/* If the page cannot be used file it away */
M
Maneesh Soni 已提交
729 730
		if (page_to_pfn(page) >
				(KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) {
731 732 733 734 735 736 737 738 739 740
			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
741 742
		if (!kimage_is_destination_range(image, addr,
						  addr + PAGE_SIZE))
743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761
			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
762 763
			 * destination page, so return it if it's
			 * gfp_flags honor the ones passed in.
764
			 */
765 766 767 768 769
			if (!(gfp_mask & __GFP_HIGHMEM) &&
			    PageHighMem(old_page)) {
				kimage_free_pages(old_page);
				continue;
			}
770 771 772 773 774 775 776 777 778 779 780
			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
781

782 783 784 785
	return page;
}

static int kimage_load_normal_segment(struct kimage *image,
M
Maneesh Soni 已提交
786
					 struct kexec_segment *segment)
787 788 789 790
{
	unsigned long maddr;
	unsigned long ubytes, mbytes;
	int result;
791
	unsigned char __user *buf;
792 793 794 795 796 797 798 799

	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
800
	if (result < 0)
801
		goto out;
M
Maneesh Soni 已提交
802 803

	while (mbytes) {
804 805 806
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
807

808
		page = kimage_alloc_page(image, GFP_HIGHUSER, maddr);
809
		if (!page) {
810 811 812
			result  = -ENOMEM;
			goto out;
		}
M
Maneesh Soni 已提交
813 814 815
		result = kimage_add_page(image, page_to_pfn(page)
								<< PAGE_SHIFT);
		if (result < 0)
816
			goto out;
M
Maneesh Soni 已提交
817

818 819
		ptr = kmap(page);
		/* Start with a clear page */
820
		clear_page(ptr);
821 822
		ptr += maddr & ~PAGE_MASK;
		mchunk = PAGE_SIZE - (maddr & ~PAGE_MASK);
M
Maneesh Soni 已提交
823
		if (mchunk > mbytes)
824
			mchunk = mbytes;
M
Maneesh Soni 已提交
825

826
		uchunk = mchunk;
M
Maneesh Soni 已提交
827
		if (uchunk > ubytes)
828
			uchunk = ubytes;
M
Maneesh Soni 已提交
829

830 831 832
		result = copy_from_user(ptr, buf, uchunk);
		kunmap(page);
		if (result) {
833
			result = -EFAULT;
834 835 836 837 838 839 840
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
		buf    += mchunk;
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
841
out:
842 843 844 845
	return result;
}

static int kimage_load_crash_segment(struct kimage *image,
M
Maneesh Soni 已提交
846
					struct kexec_segment *segment)
847 848 849 850 851 852 853 854
{
	/* 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;
	unsigned long ubytes, mbytes;
	int result;
855
	unsigned char __user *buf;
856 857 858 859 860 861

	result = 0;
	buf = segment->buf;
	ubytes = segment->bufsz;
	mbytes = segment->memsz;
	maddr = segment->mem;
M
Maneesh Soni 已提交
862
	while (mbytes) {
863 864 865
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
866

867
		page = pfn_to_page(maddr >> PAGE_SHIFT);
868
		if (!page) {
869 870 871 872 873 874
			result  = -ENOMEM;
			goto out;
		}
		ptr = kmap(page);
		ptr += maddr & ~PAGE_MASK;
		mchunk = PAGE_SIZE - (maddr & ~PAGE_MASK);
M
Maneesh Soni 已提交
875
		if (mchunk > mbytes)
876
			mchunk = mbytes;
M
Maneesh Soni 已提交
877

878 879 880 881 882 883 884
		uchunk = mchunk;
		if (uchunk > ubytes) {
			uchunk = ubytes;
			/* 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
885
		kexec_flush_icache_page(page);
886 887
		kunmap(page);
		if (result) {
888
			result = -EFAULT;
889 890 891 892 893 894 895
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
		buf    += mchunk;
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
896
out:
897 898 899 900
	return result;
}

static int kimage_load_segment(struct kimage *image,
M
Maneesh Soni 已提交
901
				struct kexec_segment *segment)
902 903
{
	int result = -ENOMEM;
M
Maneesh Soni 已提交
904 905

	switch (image->type) {
906 907 908 909 910 911 912
	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
913

914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936
	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
 *   and the copies the image to it's final destination.  And
 *   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.
 */
937 938
struct kimage *kexec_image;
struct kimage *kexec_crash_image;
939 940

static DEFINE_MUTEX(kexec_mutex);
941

942 943
SYSCALL_DEFINE4(kexec_load, unsigned long, entry, unsigned long, nr_segments,
		struct kexec_segment __user *, segments, unsigned long, flags)
944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980
{
	struct kimage **dest_image, *image;
	int result;

	/* We only trust the superuser with rebooting the system. */
	if (!capable(CAP_SYS_BOOT))
		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.
	 */
981
	if (!mutex_trylock(&kexec_mutex))
982
		return -EBUSY;
M
Maneesh Soni 已提交
983

984
	dest_image = &kexec_image;
M
Maneesh Soni 已提交
985
	if (flags & KEXEC_ON_CRASH)
986 987 988
		dest_image = &kexec_crash_image;
	if (nr_segments > 0) {
		unsigned long i;
M
Maneesh Soni 已提交
989

990
		/* Loading another kernel to reboot into */
M
Maneesh Soni 已提交
991 992 993
		if ((flags & KEXEC_ON_CRASH) == 0)
			result = kimage_normal_alloc(&image, entry,
							nr_segments, segments);
994 995 996 997 998 999
		/* 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
1000 1001
			result = kimage_crash_alloc(&image, entry,
						     nr_segments, segments);
1002
			crash_map_reserved_pages();
1003
		}
M
Maneesh Soni 已提交
1004
		if (result)
1005
			goto out;
M
Maneesh Soni 已提交
1006

H
Huang Ying 已提交
1007 1008
		if (flags & KEXEC_PRESERVE_CONTEXT)
			image->preserve_context = 1;
1009
		result = machine_kexec_prepare(image);
M
Maneesh Soni 已提交
1010
		if (result)
1011
			goto out;
M
Maneesh Soni 已提交
1012 1013

		for (i = 0; i < nr_segments; i++) {
1014
			result = kimage_load_segment(image, &image->segment[i]);
M
Maneesh Soni 已提交
1015
			if (result)
1016 1017
				goto out;
		}
1018
		kimage_terminate(image);
1019 1020
		if (flags & KEXEC_ON_CRASH)
			crash_unmap_reserved_pages();
1021 1022 1023 1024
	}
	/* Install the new kernel, and  Uninstall the old */
	image = xchg(dest_image, image);

M
Maneesh Soni 已提交
1025
out:
1026
	mutex_unlock(&kexec_mutex);
1027
	kimage_free(image);
M
Maneesh Soni 已提交
1028

1029 1030 1031
	return result;
}

1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
/*
 * 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)
{}

1044 1045
#ifdef CONFIG_COMPAT
asmlinkage long compat_sys_kexec_load(unsigned long entry,
M
Maneesh Soni 已提交
1046 1047 1048
				unsigned long nr_segments,
				struct compat_kexec_segment __user *segments,
				unsigned long flags)
1049 1050 1051 1052 1053 1054 1055 1056
{
	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
1057
	if ((flags & KEXEC_ARCH_MASK) == KEXEC_ARCH_DEFAULT)
1058 1059
		return -EINVAL;

M
Maneesh Soni 已提交
1060
	if (nr_segments > KEXEC_SEGMENT_MAX)
1061 1062 1063 1064 1065
		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
1066
		if (result)
1067 1068 1069 1070 1071 1072 1073 1074
			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
1075
		if (result)
1076 1077 1078 1079 1080 1081 1082
			return -EFAULT;
	}

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

1083
void crash_kexec(struct pt_regs *regs)
1084
{
1085
	/* Take the kexec_mutex here to prevent sys_kexec_load
1086 1087 1088 1089 1090 1091 1092
	 * 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...
	 */
1093
	if (mutex_trylock(&kexec_mutex)) {
1094
		if (kexec_crash_image) {
1095
			struct pt_regs fixed_regs;
1096

1097
			crash_setup_regs(&fixed_regs, regs);
K
Ken'ichi Ohmichi 已提交
1098
			crash_save_vmcoreinfo();
1099
			machine_crash_shutdown(&fixed_regs);
1100
			machine_kexec(kexec_crash_image);
1101
		}
1102
		mutex_unlock(&kexec_mutex);
1103 1104
	}
}
1105

1106 1107
size_t crash_get_memory_size(void)
{
1108
	size_t size = 0;
1109
	mutex_lock(&kexec_mutex);
1110
	if (crashk_res.end != crashk_res.start)
1111
		size = resource_size(&crashk_res);
1112 1113 1114 1115
	mutex_unlock(&kexec_mutex);
	return size;
}

1116 1117
void __weak crash_free_reserved_phys_range(unsigned long begin,
					   unsigned long end)
1118 1119 1120
{
	unsigned long addr;

1121 1122
	for (addr = begin; addr < end; addr += PAGE_SIZE)
		free_reserved_page(pfn_to_page(addr >> PAGE_SHIFT));
1123 1124 1125 1126 1127 1128
}

int crash_shrink_memory(unsigned long new_size)
{
	int ret = 0;
	unsigned long start, end;
1129
	unsigned long old_size;
1130
	struct resource *ram_res;
1131 1132 1133 1134 1135 1136 1137 1138 1139

	mutex_lock(&kexec_mutex);

	if (kexec_crash_image) {
		ret = -ENOENT;
		goto unlock;
	}
	start = crashk_res.start;
	end = crashk_res.end;
1140 1141 1142
	old_size = (end == 0) ? 0 : end - start + 1;
	if (new_size >= old_size) {
		ret = (new_size == old_size) ? 0 : -EINVAL;
1143 1144 1145
		goto unlock;
	}

1146 1147 1148 1149 1150 1151
	ram_res = kzalloc(sizeof(*ram_res), GFP_KERNEL);
	if (!ram_res) {
		ret = -ENOMEM;
		goto unlock;
	}

1152 1153
	start = roundup(start, KEXEC_CRASH_MEM_ALIGN);
	end = roundup(start + new_size, KEXEC_CRASH_MEM_ALIGN);
1154

1155
	crash_map_reserved_pages();
1156
	crash_free_reserved_phys_range(end, crashk_res.end);
1157

1158
	if ((start == end) && (crashk_res.parent != NULL))
1159
		release_resource(&crashk_res);
1160 1161 1162 1163 1164 1165

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

1166
	crashk_res.end = end - 1;
1167 1168

	insert_resource(&iomem_resource, ram_res);
1169
	crash_unmap_reserved_pages();
1170 1171 1172 1173 1174 1175

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

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 1206 1207 1208
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;

1209
	if ((cpu < 0) || (cpu >= nr_cpu_ids))
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223
		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
1224
	elf_core_copy_kernel_regs(&prstatus.pr_reg, regs);
1225 1226
	buf = append_elf_note(buf, KEXEC_CORE_NOTE_NAME, NT_PRSTATUS,
		      	      &prstatus, sizeof(prstatus));
1227 1228 1229
	final_note(buf);
}

1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241
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
1242

B
Bernhard Walle 已提交
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 1312 1313 1314

/*
 * 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 ? */
1315
		if (system_ram >= start && system_ram < end) {
B
Bernhard Walle 已提交
1316 1317 1318 1319 1320 1321
			*crash_size = size;
			break;
		}
	} while (*cur++ == ',');

	if (*crash_size > 0) {
1322
		while (*cur && *cur != ' ' && *cur != '@')
B
Bernhard Walle 已提交
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
			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);
1359 1360 1361 1362
	else if (*cur != ' ' && *cur != '\0') {
		pr_warning("crashkernel: unrecognized char\n");
		return -EINVAL;
	}
B
Bernhard Walle 已提交
1363 1364 1365 1366

	return 0;
}

1367 1368 1369 1370 1371 1372 1373 1374 1375
#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
1376
/*
1377 1378 1379 1380 1381
 * 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
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 1448 1449 1450
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;
}

1451
static int __init __parse_crashkernel(char *cmdline,
B
Bernhard Walle 已提交
1452 1453
			     unsigned long long system_ram,
			     unsigned long long *crash_size,
1454
			     unsigned long long *crash_base,
1455 1456
			     const char *name,
			     const char *suffix)
B
Bernhard Walle 已提交
1457 1458
{
	char	*first_colon, *first_space;
1459
	char	*ck_cmdline;
B
Bernhard Walle 已提交
1460 1461 1462 1463 1464

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

1465
	ck_cmdline = get_last_crashkernel(cmdline, name, suffix);
B
Bernhard Walle 已提交
1466 1467 1468 1469

	if (!ck_cmdline)
		return -EINVAL;

1470
	ck_cmdline += strlen(name);
B
Bernhard Walle 已提交
1471

1472 1473 1474
	if (suffix)
		return parse_crashkernel_suffix(ck_cmdline, crash_size,
				crash_base, suffix);
B
Bernhard Walle 已提交
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490
	/*
	 * 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);
	else
		return parse_crashkernel_simple(ck_cmdline, crash_size,
				crash_base);

	return 0;
}

1491 1492 1493 1494
/*
 * That function is the entry point for command line parsing and should be
 * called from the arch-specific code.
 */
1495 1496 1497 1498 1499 1500
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,
1501
					"crashkernel=", NULL);
1502
}
1503 1504 1505 1506 1507 1508 1509

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,
1510
				"crashkernel=", suffix_tbl[SUFFIX_HIGH]);
1511
}
1512 1513 1514 1515 1516 1517 1518

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,
1519
				"crashkernel=", suffix_tbl[SUFFIX_LOW]);
1520
}
B
Bernhard Walle 已提交
1521

1522
static void update_vmcoreinfo_note(void)
K
Ken'ichi Ohmichi 已提交
1523
{
1524
	u32 *buf = vmcoreinfo_note;
K
Ken'ichi Ohmichi 已提交
1525 1526 1527 1528 1529 1530 1531 1532

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

1533 1534
void crash_save_vmcoreinfo(void)
{
1535
	vmcoreinfo_append_str("CRASHTIME=%ld\n", get_seconds());
1536 1537 1538
	update_vmcoreinfo_note();
}

K
Ken'ichi Ohmichi 已提交
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570
void vmcoreinfo_append_str(const char *fmt, ...)
{
	va_list args;
	char buf[0x50];
	int r;

	va_start(args, fmt);
	r = vsnprintf(buf, sizeof(buf), fmt, args);
	va_end(args);

	if (r + vmcoreinfo_size > vmcoreinfo_max_size)
		r = vmcoreinfo_max_size - vmcoreinfo_size;

	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)
{
1571 1572
	VMCOREINFO_OSRELEASE(init_uts_ns.name.release);
	VMCOREINFO_PAGESIZE(PAGE_SIZE);
K
Ken'ichi Ohmichi 已提交
1573

1574 1575
	VMCOREINFO_SYMBOL(init_uts_ns);
	VMCOREINFO_SYMBOL(node_online_map);
1576
#ifdef CONFIG_MMU
1577
	VMCOREINFO_SYMBOL(swapper_pg_dir);
1578
#endif
1579
	VMCOREINFO_SYMBOL(_stext);
1580
	VMCOREINFO_SYMBOL(vmap_area_list);
K
Ken'ichi Ohmichi 已提交
1581 1582

#ifndef CONFIG_NEED_MULTIPLE_NODES
1583 1584
	VMCOREINFO_SYMBOL(mem_map);
	VMCOREINFO_SYMBOL(contig_page_data);
K
Ken'ichi Ohmichi 已提交
1585 1586
#endif
#ifdef CONFIG_SPARSEMEM
1587 1588
	VMCOREINFO_SYMBOL(mem_section);
	VMCOREINFO_LENGTH(mem_section, NR_SECTION_ROOTS);
1589
	VMCOREINFO_STRUCT_SIZE(mem_section);
1590
	VMCOREINFO_OFFSET(mem_section, section_mem_map);
K
Ken'ichi Ohmichi 已提交
1591
#endif
1592 1593 1594 1595 1596 1597
	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);
1598 1599 1600 1601
	VMCOREINFO_OFFSET(page, flags);
	VMCOREINFO_OFFSET(page, _count);
	VMCOREINFO_OFFSET(page, mapping);
	VMCOREINFO_OFFSET(page, lru);
1602 1603
	VMCOREINFO_OFFSET(page, _mapcount);
	VMCOREINFO_OFFSET(page, private);
1604 1605
	VMCOREINFO_OFFSET(pglist_data, node_zones);
	VMCOREINFO_OFFSET(pglist_data, nr_zones);
K
Ken'ichi Ohmichi 已提交
1606
#ifdef CONFIG_FLAT_NODE_MEM_MAP
1607
	VMCOREINFO_OFFSET(pglist_data, node_mem_map);
K
Ken'ichi Ohmichi 已提交
1608
#endif
1609 1610 1611 1612 1613 1614 1615 1616 1617
	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);
1618 1619
	VMCOREINFO_OFFSET(vmap_area, va_start);
	VMCOREINFO_OFFSET(vmap_area, list);
1620
	VMCOREINFO_LENGTH(zone.free_area, MAX_ORDER);
1621
	log_buf_kexec_setup();
1622
	VMCOREINFO_LENGTH(free_area.free_list, MIGRATE_TYPES);
1623
	VMCOREINFO_NUMBER(NR_FREE_PAGES);
1624 1625 1626
	VMCOREINFO_NUMBER(PG_lru);
	VMCOREINFO_NUMBER(PG_private);
	VMCOREINFO_NUMBER(PG_swapcache);
1627
	VMCOREINFO_NUMBER(PG_slab);
1628 1629 1630
#ifdef CONFIG_MEMORY_FAILURE
	VMCOREINFO_NUMBER(PG_hwpoison);
#endif
1631
	VMCOREINFO_NUMBER(PAGE_BUDDY_MAPCOUNT_VALUE);
K
Ken'ichi Ohmichi 已提交
1632 1633

	arch_crash_save_vmcoreinfo();
1634
	update_vmcoreinfo_note();
K
Ken'ichi Ohmichi 已提交
1635 1636 1637 1638 1639

	return 0;
}

module_init(crash_save_vmcoreinfo_init)
H
Huang Ying 已提交
1640

1641 1642 1643
/*
 * Move into place and start executing a preloaded standalone
 * executable.  If nothing was preloaded return an error.
H
Huang Ying 已提交
1644 1645 1646 1647 1648
 */
int kernel_kexec(void)
{
	int error = 0;

1649
	if (!mutex_trylock(&kexec_mutex))
H
Huang Ying 已提交
1650 1651 1652 1653 1654 1655 1656
		return -EBUSY;
	if (!kexec_image) {
		error = -EINVAL;
		goto Unlock;
	}

#ifdef CONFIG_KEXEC_JUMP
1657
	if (kexec_image->preserve_context) {
1658
		lock_system_sleep();
1659 1660 1661 1662 1663 1664 1665
		pm_prepare_console();
		error = freeze_processes();
		if (error) {
			error = -EBUSY;
			goto Restore_console;
		}
		suspend_console();
1666
		error = dpm_suspend_start(PMSG_FREEZE);
1667 1668
		if (error)
			goto Resume_console;
1669
		/* At this point, dpm_suspend_start() has been called,
1670 1671
		 * but *not* dpm_suspend_end(). We *must* call
		 * dpm_suspend_end() now.  Otherwise, drivers for
1672 1673 1674 1675
		 * some devices (e.g. interrupt controllers) become
		 * desynchronized with the actual state of the
		 * hardware at resume time, and evil weirdness ensues.
		 */
1676
		error = dpm_suspend_end(PMSG_FREEZE);
1677
		if (error)
1678 1679 1680 1681
			goto Resume_devices;
		error = disable_nonboot_cpus();
		if (error)
			goto Enable_cpus;
1682
		local_irq_disable();
1683
		error = syscore_suspend();
1684
		if (error)
1685
			goto Enable_irqs;
1686
	} else
H
Huang Ying 已提交
1687
#endif
1688
	{
1689
		kernel_restart_prepare(NULL);
H
Huang Ying 已提交
1690 1691 1692 1693 1694 1695 1696
		printk(KERN_EMERG "Starting new kernel\n");
		machine_shutdown();
	}

	machine_kexec(kexec_image);

#ifdef CONFIG_KEXEC_JUMP
1697
	if (kexec_image->preserve_context) {
1698
		syscore_resume();
1699
 Enable_irqs:
H
Huang Ying 已提交
1700
		local_irq_enable();
1701
 Enable_cpus:
1702
		enable_nonboot_cpus();
1703
		dpm_resume_start(PMSG_RESTORE);
1704
 Resume_devices:
1705
		dpm_resume_end(PMSG_RESTORE);
1706 1707 1708 1709 1710
 Resume_console:
		resume_console();
		thaw_processes();
 Restore_console:
		pm_restore_console();
1711
		unlock_system_sleep();
H
Huang Ying 已提交
1712
	}
1713
#endif
H
Huang Ying 已提交
1714 1715

 Unlock:
1716
	mutex_unlock(&kexec_mutex);
H
Huang Ying 已提交
1717 1718
	return error;
}