kexec.c 27.5 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
/*
 * 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.
 */

#include <linux/mm.h>
#include <linux/file.h>
#include <linux/slab.h>
#include <linux/fs.h>
#include <linux/kexec.h>
#include <linux/spinlock.h>
#include <linux/list.h>
#include <linux/highmem.h>
#include <linux/syscalls.h>
#include <linux/reboot.h>
#include <linux/syscalls.h>
#include <linux/ioport.h>
21 22
#include <linux/hardirq.h>

23 24 25 26 27 28
#include <asm/page.h>
#include <asm/uaccess.h>
#include <asm/io.h>
#include <asm/system.h>
#include <asm/semaphore.h>

29 30 31
/* Per cpu memory for storing cpu states in case of system crash. */
note_buf_t* crash_notes;

32 33 34 35 36 37 38 39
/* 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
};

40 41 42 43 44 45 46
int kexec_should_crash(struct task_struct *p)
{
	if (in_interrupt() || !p->pid || p->pid == 1 || panic_on_oops)
		return 1;
	return 0;
}

47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92
/*
 * 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
 * is given by KEXEC_CONTROL_CODE_SIZE.  In the best case only a single
 * 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
93 94 95
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
96
				       gfp_t gfp_mask,
M
Maneesh Soni 已提交
97
				       unsigned long dest);
98 99

static int do_kimage_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
100 101
	                    unsigned long nr_segments,
                            struct kexec_segment __user *segments)
102 103 104 105 106 107 108 109 110
{
	size_t segment_bytes;
	struct kimage *image;
	unsigned long i;
	int result;

	/* Allocate a controlling structure */
	result = -ENOMEM;
	image = kmalloc(sizeof(*image), GFP_KERNEL);
M
Maneesh Soni 已提交
111
	if (!image)
112
		goto out;
M
Maneesh Soni 已提交
113

114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153
	memset(image, 0, sizeof(*image));
	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);

	/* Initialize the list of unuseable pages */
	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);
	if (result)
		goto out;

	/*
	 * 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
	 * the destination addreses are page aligned.  Too many
	 * 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
154

155 156 157 158 159 160 161 162 163 164 165 166 167 168
		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
169
	for (i = 0; i < nr_segments; i++) {
170 171
		unsigned long mstart, mend;
		unsigned long j;
M
Maneesh Soni 已提交
172

173 174
		mstart = image->segment[i].mem;
		mend   = mstart + image->segment[i].memsz;
M
Maneesh Soni 已提交
175
		for (j = 0; j < i; j++) {
176 177 178 179 180 181 182 183 184 185 186 187 188 189 190
			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
191
	for (i = 0; i < nr_segments; i++) {
192 193 194 195 196
		if (image->segment[i].bufsz > image->segment[i].memsz)
			goto out;
	}

	result = 0;
M
Maneesh Soni 已提交
197 198
out:
	if (result == 0)
199
		*rimage = image;
M
Maneesh Soni 已提交
200
	else
201
		kfree(image);
M
Maneesh Soni 已提交
202

203 204 205 206 207
	return result;

}

static int kimage_normal_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
208 209
				unsigned long nr_segments,
				struct kexec_segment __user *segments)
210 211 212 213 214 215 216
{
	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
217
	if (result)
218
		goto out;
M
Maneesh Soni 已提交
219

220 221 222 223 224 225 226 227 228
	*rimage = image;

	/*
	 * 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,
M
Maneesh Soni 已提交
229
					   get_order(KEXEC_CONTROL_CODE_SIZE));
230 231 232 233 234 235 236
	if (!image->control_code_page) {
		printk(KERN_ERR "Could not allocate control_code_buffer\n");
		goto out;
	}

	result = 0;
 out:
M
Maneesh Soni 已提交
237
	if (result == 0)
238
		*rimage = image;
M
Maneesh Soni 已提交
239
	else
240
		kfree(image);
M
Maneesh Soni 已提交
241

242 243 244 245
	return result;
}

static int kimage_crash_alloc(struct kimage **rimage, unsigned long entry,
M
Maneesh Soni 已提交
246
				unsigned long nr_segments,
247
				struct kexec_segment __user *segments)
248 249 250 251 252 253 254 255 256 257 258 259 260 261
{
	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
262
	if (result)
263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282
		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
283

284
		mstart = image->segment[i].mem;
285
		mend = mstart + image->segment[i].memsz - 1;
286 287 288 289 290 291 292 293 294 295 296 297
		/* Ensure we are within the crash kernel limits */
		if ((mstart < crashk_res.start) || (mend > crashk_res.end))
			goto out;
	}

	/*
	 * 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,
M
Maneesh Soni 已提交
298
					   get_order(KEXEC_CONTROL_CODE_SIZE));
299 300 301 302 303 304
	if (!image->control_code_page) {
		printk(KERN_ERR "Could not allocate control_code_buffer\n");
		goto out;
	}

	result = 0;
M
Maneesh Soni 已提交
305 306
out:
	if (result == 0)
307
		*rimage = image;
M
Maneesh Soni 已提交
308
	else
309
		kfree(image);
M
Maneesh Soni 已提交
310

311 312 313
	return result;
}

M
Maneesh Soni 已提交
314 315 316
static int kimage_is_destination_range(struct kimage *image,
					unsigned long start,
					unsigned long end)
317 318 319 320 321
{
	unsigned long i;

	for (i = 0; i < image->nr_segments; i++) {
		unsigned long mstart, mend;
M
Maneesh Soni 已提交
322

323
		mstart = image->segment[i].mem;
M
Maneesh Soni 已提交
324 325
		mend = mstart + image->segment[i].memsz;
		if ((end > mstart) && (start < mend))
326 327
			return 1;
	}
M
Maneesh Soni 已提交
328

329 330 331
	return 0;
}

A
Al Viro 已提交
332
static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order)
333 334
{
	struct page *pages;
M
Maneesh Soni 已提交
335

336 337 338 339
	pages = alloc_pages(gfp_mask, order);
	if (pages) {
		unsigned int count, i;
		pages->mapping = NULL;
H
Hugh Dickins 已提交
340
		set_page_private(pages, order);
341
		count = 1 << order;
M
Maneesh Soni 已提交
342
		for (i = 0; i < count; i++)
343 344
			SetPageReserved(pages + i);
	}
M
Maneesh Soni 已提交
345

346 347 348 349 350 351
	return pages;
}

static void kimage_free_pages(struct page *page)
{
	unsigned int order, count, i;
M
Maneesh Soni 已提交
352

H
Hugh Dickins 已提交
353
	order = page_private(page);
354
	count = 1 << order;
M
Maneesh Soni 已提交
355
	for (i = 0; i < count; i++)
356 357 358 359 360 361 362
		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
363

364 365 366 367 368 369 370 371 372
	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
373 374
static struct page *kimage_alloc_normal_control_pages(struct kimage *image,
							unsigned int order)
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
{
	/* 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
401

402 403 404 405 406 407 408 409
		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
410
			      kimage_is_destination_range(image, addr, eaddr)) {
411 412 413
			list_add(&pages->lru, &extra_pages);
			pages = NULL;
		}
M
Maneesh Soni 已提交
414 415
	} while (!pages);

416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435
	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
	 * page allocations, and add everyting to image->dest_pages.
	 *
	 * For now it is simpler to just free the pages.
	 */
	kimage_free_page_list(&extra_pages);

M
Maneesh Soni 已提交
436
	return pages;
437 438
}

M
Maneesh Soni 已提交
439 440
static struct page *kimage_alloc_crash_control_pages(struct kimage *image,
						      unsigned int order)
441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
{
	/* 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
465

466 467 468 469
	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
470
	while (hole_end <= crashk_res.end) {
471
		unsigned long i;
M
Maneesh Soni 已提交
472 473

		if (hole_end > KEXEC_CONTROL_MEMORY_LIMIT)
474
			break;
M
Maneesh Soni 已提交
475
		if (hole_end > crashk_res.end)
476 477
			break;
		/* See if I overlap any of the segments */
M
Maneesh Soni 已提交
478
		for (i = 0; i < image->nr_segments; i++) {
479
			unsigned long mstart, mend;
M
Maneesh Soni 已提交
480

481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
			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
496
	if (pages)
497
		image->control_page = hole_end;
M
Maneesh Soni 已提交
498

499 500 501 502
	return pages;
}


M
Maneesh Soni 已提交
503 504
struct page *kimage_alloc_control_pages(struct kimage *image,
					 unsigned int order)
505 506
{
	struct page *pages = NULL;
M
Maneesh Soni 已提交
507 508

	switch (image->type) {
509 510 511 512 513 514 515
	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
516

517 518 519 520 521
	return pages;
}

static int kimage_add_entry(struct kimage *image, kimage_entry_t entry)
{
M
Maneesh Soni 已提交
522
	if (*image->entry != 0)
523
		image->entry++;
M
Maneesh Soni 已提交
524

525 526 527
	if (image->entry == image->last_entry) {
		kimage_entry_t *ind_page;
		struct page *page;
M
Maneesh Soni 已提交
528

529
		page = kimage_alloc_page(image, GFP_KERNEL, KIMAGE_NO_DEST);
M
Maneesh Soni 已提交
530
		if (!page)
531
			return -ENOMEM;
M
Maneesh Soni 已提交
532

533 534 535
		ind_page = page_address(page);
		*image->entry = virt_to_phys(ind_page) | IND_INDIRECTION;
		image->entry = ind_page;
M
Maneesh Soni 已提交
536 537
		image->last_entry = ind_page +
				      ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1);
538 539 540 541
	}
	*image->entry = entry;
	image->entry++;
	*image->entry = 0;
M
Maneesh Soni 已提交
542

543 544 545
	return 0;
}

M
Maneesh Soni 已提交
546 547
static int kimage_set_destination(struct kimage *image,
				   unsigned long destination)
548 549 550 551 552
{
	int result;

	destination &= PAGE_MASK;
	result = kimage_add_entry(image, destination | IND_DESTINATION);
M
Maneesh Soni 已提交
553
	if (result == 0)
554
		image->destination = destination;
M
Maneesh Soni 已提交
555

556 557 558 559 560 561 562 563 564 565
	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
566
	if (result == 0)
567
		image->destination += PAGE_SIZE;
M
Maneesh Soni 已提交
568

569 570 571 572 573 574 575 576 577 578 579 580 581 582 583
	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);

	/* Walk through and free any unuseable pages I have cached */
	kimage_free_page_list(&image->unuseable_pages);

}
static int kimage_terminate(struct kimage *image)
{
M
Maneesh Soni 已提交
584
	if (*image->entry != 0)
585
		image->entry++;
M
Maneesh Soni 已提交
586

587
	*image->entry = IND_DONE;
M
Maneesh Soni 已提交
588

589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611
	return 0;
}

#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
612

613 614 615 616
	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
617
			if (ind & IND_INDIRECTION)
618 619 620 621 622 623
				kimage_free_entry(ind);
			/* Save this indirection page until we are
			 * done with it.
			 */
			ind = entry;
		}
M
Maneesh Soni 已提交
624
		else if (entry & IND_SOURCE)
625 626 627
			kimage_free_entry(entry);
	}
	/* Free the final indirection page */
M
Maneesh Soni 已提交
628
	if (ind & IND_INDIRECTION)
629 630 631 632 633 634 635 636 637 638
		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
639 640
static kimage_entry_t *kimage_dst_used(struct kimage *image,
					unsigned long page)
641 642 643 644 645
{
	kimage_entry_t *ptr, entry;
	unsigned long destination = 0;

	for_each_kimage_entry(image, ptr, entry) {
M
Maneesh Soni 已提交
646
		if (entry & IND_DESTINATION)
647 648
			destination = entry & PAGE_MASK;
		else if (entry & IND_SOURCE) {
M
Maneesh Soni 已提交
649
			if (page == destination)
650 651 652 653
				return ptr;
			destination += PAGE_SIZE;
		}
	}
M
Maneesh Soni 已提交
654

655
	return NULL;
656 657
}

M
Maneesh Soni 已提交
658
static struct page *kimage_alloc_page(struct kimage *image,
A
Al Viro 已提交
659
					gfp_t gfp_mask,
M
Maneesh Soni 已提交
660
					unsigned long destination)
661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699
{
	/*
	 * 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
700
		if (!page)
701
			return NULL;
702
		/* If the page cannot be used file it away */
M
Maneesh Soni 已提交
703 704
		if (page_to_pfn(page) >
				(KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) {
705 706 707 708 709 710 711 712 713 714
			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
715 716
		if (!kimage_is_destination_range(image, addr,
						  addr + PAGE_SIZE))
717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748
			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
			 * destination page, so return it.
			 */
			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
749

750 751 752 753
	return page;
}

static int kimage_load_normal_segment(struct kimage *image,
M
Maneesh Soni 已提交
754
					 struct kexec_segment *segment)
755 756 757 758
{
	unsigned long maddr;
	unsigned long ubytes, mbytes;
	int result;
759
	unsigned char __user *buf;
760 761 762 763 764 765 766 767

	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
768
	if (result < 0)
769
		goto out;
M
Maneesh Soni 已提交
770 771

	while (mbytes) {
772 773 774
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
775

776 777 778 779 780
		page = kimage_alloc_page(image, GFP_HIGHUSER, maddr);
		if (page == 0) {
			result  = -ENOMEM;
			goto out;
		}
M
Maneesh Soni 已提交
781 782 783
		result = kimage_add_page(image, page_to_pfn(page)
								<< PAGE_SHIFT);
		if (result < 0)
784
			goto out;
M
Maneesh Soni 已提交
785

786 787 788 789 790
		ptr = kmap(page);
		/* Start with a clear page */
		memset(ptr, 0, PAGE_SIZE);
		ptr += maddr & ~PAGE_MASK;
		mchunk = PAGE_SIZE - (maddr & ~PAGE_MASK);
M
Maneesh Soni 已提交
791
		if (mchunk > mbytes)
792
			mchunk = mbytes;
M
Maneesh Soni 已提交
793

794
		uchunk = mchunk;
M
Maneesh Soni 已提交
795
		if (uchunk > ubytes)
796
			uchunk = ubytes;
M
Maneesh Soni 已提交
797

798 799 800 801 802 803 804 805 806 807 808
		result = copy_from_user(ptr, buf, uchunk);
		kunmap(page);
		if (result) {
			result = (result < 0) ? result : -EIO;
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
		buf    += mchunk;
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
809
out:
810 811 812 813
	return result;
}

static int kimage_load_crash_segment(struct kimage *image,
M
Maneesh Soni 已提交
814
					struct kexec_segment *segment)
815 816 817 818 819 820 821 822
{
	/* 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;
823
	unsigned char __user *buf;
824 825 826 827 828 829

	result = 0;
	buf = segment->buf;
	ubytes = segment->bufsz;
	mbytes = segment->memsz;
	maddr = segment->mem;
M
Maneesh Soni 已提交
830
	while (mbytes) {
831 832 833
		struct page *page;
		char *ptr;
		size_t uchunk, mchunk;
M
Maneesh Soni 已提交
834

835 836 837 838 839 840 841 842
		page = pfn_to_page(maddr >> PAGE_SHIFT);
		if (page == 0) {
			result  = -ENOMEM;
			goto out;
		}
		ptr = kmap(page);
		ptr += maddr & ~PAGE_MASK;
		mchunk = PAGE_SIZE - (maddr & ~PAGE_MASK);
M
Maneesh Soni 已提交
843
		if (mchunk > mbytes)
844
			mchunk = mbytes;
M
Maneesh Soni 已提交
845

846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862
		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);
		kunmap(page);
		if (result) {
			result = (result < 0) ? result : -EIO;
			goto out;
		}
		ubytes -= uchunk;
		maddr  += mchunk;
		buf    += mchunk;
		mbytes -= mchunk;
	}
M
Maneesh Soni 已提交
863
out:
864 865 866 867
	return result;
}

static int kimage_load_segment(struct kimage *image,
M
Maneesh Soni 已提交
868
				struct kexec_segment *segment)
869 870
{
	int result = -ENOMEM;
M
Maneesh Soni 已提交
871 872

	switch (image->type) {
873 874 875 876 877 878 879
	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
880

881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912
	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.
 */
struct kimage *kexec_image = NULL;
static struct kimage *kexec_crash_image = NULL;
/*
 * A home grown binary mutex.
 * Nothing can wait so this mutex is safe to use
 * in interrupt context :)
 */
static int kexec_lock = 0;

M
Maneesh Soni 已提交
913 914 915
asmlinkage long sys_kexec_load(unsigned long entry, unsigned long nr_segments,
				struct kexec_segment __user *segments,
				unsigned long flags)
916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954
{
	struct kimage **dest_image, *image;
	int locked;
	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.
	 */
	locked = xchg(&kexec_lock, 1);
M
Maneesh Soni 已提交
955
	if (locked)
956
		return -EBUSY;
M
Maneesh Soni 已提交
957

958
	dest_image = &kexec_image;
M
Maneesh Soni 已提交
959
	if (flags & KEXEC_ON_CRASH)
960 961 962
		dest_image = &kexec_crash_image;
	if (nr_segments > 0) {
		unsigned long i;
M
Maneesh Soni 已提交
963

964
		/* Loading another kernel to reboot into */
M
Maneesh Soni 已提交
965 966 967
		if ((flags & KEXEC_ON_CRASH) == 0)
			result = kimage_normal_alloc(&image, entry,
							nr_segments, segments);
968 969 970 971 972 973
		/* 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
974 975
			result = kimage_crash_alloc(&image, entry,
						     nr_segments, segments);
976
		}
M
Maneesh Soni 已提交
977
		if (result)
978
			goto out;
M
Maneesh Soni 已提交
979

980
		result = machine_kexec_prepare(image);
M
Maneesh Soni 已提交
981
		if (result)
982
			goto out;
M
Maneesh Soni 已提交
983 984

		for (i = 0; i < nr_segments; i++) {
985
			result = kimage_load_segment(image, &image->segment[i]);
M
Maneesh Soni 已提交
986
			if (result)
987 988 989
				goto out;
		}
		result = kimage_terminate(image);
M
Maneesh Soni 已提交
990
		if (result)
991 992 993 994 995
			goto out;
	}
	/* Install the new kernel, and  Uninstall the old */
	image = xchg(dest_image, image);

M
Maneesh Soni 已提交
996
out:
997 998
	xchg(&kexec_lock, 0); /* Release the mutex */
	kimage_free(image);
M
Maneesh Soni 已提交
999

1000 1001 1002 1003 1004
	return result;
}

#ifdef CONFIG_COMPAT
asmlinkage long compat_sys_kexec_load(unsigned long entry,
M
Maneesh Soni 已提交
1005 1006 1007
				unsigned long nr_segments,
				struct compat_kexec_segment __user *segments,
				unsigned long flags)
1008 1009 1010 1011 1012 1013 1014 1015
{
	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
1016
	if ((flags & KEXEC_ARCH_MASK) == KEXEC_ARCH_DEFAULT)
1017 1018
		return -EINVAL;

M
Maneesh Soni 已提交
1019
	if (nr_segments > KEXEC_SEGMENT_MAX)
1020 1021 1022 1023 1024
		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
1025
		if (result)
1026 1027 1028 1029 1030 1031 1032 1033
			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
1034
		if (result)
1035 1036 1037 1038 1039 1040 1041
			return -EFAULT;
	}

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

1042
void crash_kexec(struct pt_regs *regs)
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059
{
	struct kimage *image;
	int locked;


	/* Take the kexec_lock here to prevent sys_kexec_load
	 * 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...
	 */
	locked = xchg(&kexec_lock, 1);
	if (!locked) {
		image = xchg(&kexec_crash_image, NULL);
		if (image) {
1060
			machine_crash_shutdown(regs);
1061 1062 1063 1064 1065
			machine_kexec(image);
		}
		xchg(&kexec_lock, 0);
	}
}
1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078

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)