binfmt_elf.c 54.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
/*
 * linux/fs/binfmt_elf.c
 *
 * These are the functions used to load ELF format executables as used
 * on SVr4 machines.  Information on the format may be found in the book
 * "UNIX SYSTEM V RELEASE 4 Programmers Guide: Ansi C and Programming Support
 * Tools".
 *
 * Copyright 1993, 1994: Eric Youngdale (ericy@cais.com).
 */

#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/fs.h>
#include <linux/mm.h>
#include <linux/mman.h>
#include <linux/errno.h>
#include <linux/signal.h>
#include <linux/binfmts.h>
#include <linux/string.h>
#include <linux/file.h>
#include <linux/slab.h>
#include <linux/personality.h>
#include <linux/elfcore.h>
#include <linux/init.h>
#include <linux/highuid.h>
#include <linux/compiler.h>
#include <linux/highmem.h>
#include <linux/pagemap.h>
#include <linux/security.h>
#include <linux/random.h>
32
#include <linux/elf.h>
33
#include <linux/utsname.h>
L
Linus Torvalds 已提交
34 35 36 37
#include <asm/uaccess.h>
#include <asm/param.h>
#include <asm/page.h>

38 39
static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs);
static int load_elf_library(struct file *);
40 41
static unsigned long elf_map(struct file *, unsigned long, struct elf_phdr *,
				int, int, unsigned long);
L
Linus Torvalds 已提交
42 43 44 45 46

/*
 * If we don't support core dumping, then supply a NULL so we
 * don't even try.
 */
47
#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE)
48
static int elf_core_dump(long signr, struct pt_regs *regs, struct file *file, unsigned long limit);
L
Linus Torvalds 已提交
49 50 51 52 53
#else
#define elf_core_dump	NULL
#endif

#if ELF_EXEC_PAGESIZE > PAGE_SIZE
54
#define ELF_MIN_ALIGN	ELF_EXEC_PAGESIZE
L
Linus Torvalds 已提交
55
#else
56
#define ELF_MIN_ALIGN	PAGE_SIZE
L
Linus Torvalds 已提交
57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
#endif

#ifndef ELF_CORE_EFLAGS
#define ELF_CORE_EFLAGS	0
#endif

#define ELF_PAGESTART(_v) ((_v) & ~(unsigned long)(ELF_MIN_ALIGN-1))
#define ELF_PAGEOFFSET(_v) ((_v) & (ELF_MIN_ALIGN-1))
#define ELF_PAGEALIGN(_v) (((_v) + ELF_MIN_ALIGN - 1) & ~(ELF_MIN_ALIGN - 1))

static struct linux_binfmt elf_format = {
		.module		= THIS_MODULE,
		.load_binary	= load_elf_binary,
		.load_shlib	= load_elf_library,
		.core_dump	= elf_core_dump,
72 73
		.min_coredump	= ELF_EXEC_PAGESIZE,
		.hasvdso	= 1
L
Linus Torvalds 已提交
74 75
};

A
Andrew Morton 已提交
76
#define BAD_ADDR(x) ((unsigned long)(x) >= TASK_SIZE)
L
Linus Torvalds 已提交
77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96

static int set_brk(unsigned long start, unsigned long end)
{
	start = ELF_PAGEALIGN(start);
	end = ELF_PAGEALIGN(end);
	if (end > start) {
		unsigned long addr;
		down_write(&current->mm->mmap_sem);
		addr = do_brk(start, end - start);
		up_write(&current->mm->mmap_sem);
		if (BAD_ADDR(addr))
			return addr;
	}
	current->mm->start_brk = current->mm->brk = end;
	return 0;
}

/* We need to explicitly zero any fractional pages
   after the data section (i.e. bss).  This would
   contain the junk from the file that should not
97 98
   be in memory
 */
L
Linus Torvalds 已提交
99 100 101 102 103 104 105 106 107 108 109 110 111
static int padzero(unsigned long elf_bss)
{
	unsigned long nbyte;

	nbyte = ELF_PAGEOFFSET(elf_bss);
	if (nbyte) {
		nbyte = ELF_MIN_ALIGN - nbyte;
		if (clear_user((void __user *) elf_bss, nbyte))
			return -EFAULT;
	}
	return 0;
}

O
Ohad Ben-Cohen 已提交
112
/* Let's use some macros to make this stack manipulation a little clearer */
L
Linus Torvalds 已提交
113 114 115 116
#ifdef CONFIG_STACK_GROWSUP
#define STACK_ADD(sp, items) ((elf_addr_t __user *)(sp) + (items))
#define STACK_ROUND(sp, items) \
	((15 + (unsigned long) ((sp) + (items))) &~ 15UL)
117 118 119
#define STACK_ALLOC(sp, len) ({ \
	elf_addr_t __user *old_sp = (elf_addr_t __user *)sp; sp += len; \
	old_sp; })
L
Linus Torvalds 已提交
120 121 122 123 124 125 126
#else
#define STACK_ADD(sp, items) ((elf_addr_t __user *)(sp) - (items))
#define STACK_ROUND(sp, items) \
	(((unsigned long) (sp - items)) &~ 15UL)
#define STACK_ALLOC(sp, len) ({ sp -= len ; sp; })
#endif

127 128 129 130 131 132 133 134 135
#ifndef ELF_BASE_PLATFORM
/*
 * AT_BASE_PLATFORM indicates the "real" hardware/microarchitecture.
 * If the arch defines ELF_BASE_PLATFORM (in asm/elf.h), the value
 * will be copied to the user stack in the same manner as AT_PLATFORM.
 */
#define ELF_BASE_PLATFORM NULL
#endif

L
Linus Torvalds 已提交
136
static int
137
create_elf_tables(struct linux_binprm *bprm, struct elfhdr *exec,
138
		unsigned long load_addr, unsigned long interp_load_addr)
L
Linus Torvalds 已提交
139 140 141 142 143 144 145 146
{
	unsigned long p = bprm->p;
	int argc = bprm->argc;
	int envc = bprm->envc;
	elf_addr_t __user *argv;
	elf_addr_t __user *envp;
	elf_addr_t __user *sp;
	elf_addr_t __user *u_platform;
147
	elf_addr_t __user *u_base_platform;
148
	elf_addr_t __user *u_rand_bytes;
L
Linus Torvalds 已提交
149
	const char *k_platform = ELF_PLATFORM;
150
	const char *k_base_platform = ELF_BASE_PLATFORM;
151
	unsigned char k_rand_bytes[16];
L
Linus Torvalds 已提交
152 153 154
	int items;
	elf_addr_t *elf_info;
	int ei_index = 0;
155
	const struct cred *cred = current_cred();
156
	struct vm_area_struct *vma;
L
Linus Torvalds 已提交
157

158 159 160 161 162 163 164 165
	/*
	 * In some cases (e.g. Hyper-Threading), we want to avoid L1
	 * evictions by the processes running on the same package. One
	 * thing we can do is to shuffle the initial stack for them.
	 */

	p = arch_align_stack(p);

L
Linus Torvalds 已提交
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180
	/*
	 * If this architecture has a platform capability string, copy it
	 * to userspace.  In some cases (Sparc), this info is impossible
	 * for userspace to get any other way, in others (i386) it is
	 * merely difficult.
	 */
	u_platform = NULL;
	if (k_platform) {
		size_t len = strlen(k_platform) + 1;

		u_platform = (elf_addr_t __user *)STACK_ALLOC(p, len);
		if (__copy_to_user(u_platform, k_platform, len))
			return -EFAULT;
	}

181 182 183 184 185 186 187 188 189 190 191 192 193
	/*
	 * If this architecture has a "base" platform capability
	 * string, copy it to userspace.
	 */
	u_base_platform = NULL;
	if (k_base_platform) {
		size_t len = strlen(k_base_platform) + 1;

		u_base_platform = (elf_addr_t __user *)STACK_ALLOC(p, len);
		if (__copy_to_user(u_base_platform, k_base_platform, len))
			return -EFAULT;
	}

194 195 196 197 198 199 200 201 202
	/*
	 * Generate 16 random bytes for userspace PRNG seeding.
	 */
	get_random_bytes(k_rand_bytes, sizeof(k_rand_bytes));
	u_rand_bytes = (elf_addr_t __user *)
		       STACK_ALLOC(p, sizeof(k_rand_bytes));
	if (__copy_to_user(u_rand_bytes, k_rand_bytes, sizeof(k_rand_bytes)))
		return -EFAULT;

L
Linus Torvalds 已提交
203
	/* Create the ELF interpreter info */
204
	elf_info = (elf_addr_t *)current->mm->saved_auxv;
205
	/* update AT_VECTOR_SIZE_BASE if the number of NEW_AUX_ENT() changes */
L
Linus Torvalds 已提交
206
#define NEW_AUX_ENT(id, val) \
207
	do { \
208 209
		elf_info[ei_index++] = id; \
		elf_info[ei_index++] = val; \
210
	} while (0)
L
Linus Torvalds 已提交
211 212 213 214 215

#ifdef ARCH_DLINFO
	/* 
	 * ARCH_DLINFO must come first so PPC can do its special alignment of
	 * AUXV.
216 217
	 * update AT_VECTOR_SIZE_ARCH if the number of NEW_AUX_ENT() in
	 * ARCH_DLINFO changes
L
Linus Torvalds 已提交
218 219 220 221 222 223 224
	 */
	ARCH_DLINFO;
#endif
	NEW_AUX_ENT(AT_HWCAP, ELF_HWCAP);
	NEW_AUX_ENT(AT_PAGESZ, ELF_EXEC_PAGESIZE);
	NEW_AUX_ENT(AT_CLKTCK, CLOCKS_PER_SEC);
	NEW_AUX_ENT(AT_PHDR, load_addr + exec->e_phoff);
225
	NEW_AUX_ENT(AT_PHENT, sizeof(struct elf_phdr));
L
Linus Torvalds 已提交
226 227 228 229
	NEW_AUX_ENT(AT_PHNUM, exec->e_phnum);
	NEW_AUX_ENT(AT_BASE, interp_load_addr);
	NEW_AUX_ENT(AT_FLAGS, 0);
	NEW_AUX_ENT(AT_ENTRY, exec->e_entry);
230 231 232 233
	NEW_AUX_ENT(AT_UID, cred->uid);
	NEW_AUX_ENT(AT_EUID, cred->euid);
	NEW_AUX_ENT(AT_GID, cred->gid);
	NEW_AUX_ENT(AT_EGID, cred->egid);
234
 	NEW_AUX_ENT(AT_SECURE, security_bprm_secureexec(bprm));
235
	NEW_AUX_ENT(AT_RANDOM, (elf_addr_t)(unsigned long)u_rand_bytes);
236
	NEW_AUX_ENT(AT_EXECFN, bprm->exec);
L
Linus Torvalds 已提交
237
	if (k_platform) {
238
		NEW_AUX_ENT(AT_PLATFORM,
239
			    (elf_addr_t)(unsigned long)u_platform);
L
Linus Torvalds 已提交
240
	}
241 242 243 244
	if (k_base_platform) {
		NEW_AUX_ENT(AT_BASE_PLATFORM,
			    (elf_addr_t)(unsigned long)u_base_platform);
	}
L
Linus Torvalds 已提交
245
	if (bprm->interp_flags & BINPRM_FLAGS_EXECFD) {
246
		NEW_AUX_ENT(AT_EXECFD, bprm->interp_data);
L
Linus Torvalds 已提交
247 248 249 250 251 252 253 254 255 256 257
	}
#undef NEW_AUX_ENT
	/* AT_NULL is zero; clear the rest too */
	memset(&elf_info[ei_index], 0,
	       sizeof current->mm->saved_auxv - ei_index * sizeof elf_info[0]);

	/* And advance past the AT_NULL entry.  */
	ei_index += 2;

	sp = STACK_ADD(p, ei_index);

258
	items = (argc + 1) + (envc + 1) + 1;
L
Linus Torvalds 已提交
259 260 261 262 263
	bprm->p = STACK_ROUND(sp, items);

	/* Point sp at the lowest address on the stack */
#ifdef CONFIG_STACK_GROWSUP
	sp = (elf_addr_t __user *)bprm->p - items - ei_index;
264
	bprm->exec = (unsigned long)sp; /* XXX: PARISC HACK */
L
Linus Torvalds 已提交
265 266 267 268
#else
	sp = (elf_addr_t __user *)bprm->p;
#endif

269 270 271 272 273 274 275 276 277

	/*
	 * Grow the stack manually; some architectures have a limit on how
	 * far ahead a user-space access may be in order to grow the stack.
	 */
	vma = find_extend_vma(current->mm, bprm->p);
	if (!vma)
		return -EFAULT;

L
Linus Torvalds 已提交
278 279 280
	/* Now, let's put argc (and argv, envp if appropriate) on the stack */
	if (__put_user(argc, sp++))
		return -EFAULT;
281 282
	argv = sp;
	envp = argv + argc + 1;
L
Linus Torvalds 已提交
283 284

	/* Populate argv and envp */
285
	p = current->mm->arg_end = current->mm->arg_start;
L
Linus Torvalds 已提交
286 287
	while (argc-- > 0) {
		size_t len;
288 289
		if (__put_user((elf_addr_t)p, argv++))
			return -EFAULT;
290 291
		len = strnlen_user((void __user *)p, MAX_ARG_STRLEN);
		if (!len || len > MAX_ARG_STRLEN)
292
			return -EINVAL;
L
Linus Torvalds 已提交
293 294 295 296 297 298 299
		p += len;
	}
	if (__put_user(0, argv))
		return -EFAULT;
	current->mm->arg_end = current->mm->env_start = p;
	while (envc-- > 0) {
		size_t len;
300 301
		if (__put_user((elf_addr_t)p, envp++))
			return -EFAULT;
302 303
		len = strnlen_user((void __user *)p, MAX_ARG_STRLEN);
		if (!len || len > MAX_ARG_STRLEN)
304
			return -EINVAL;
L
Linus Torvalds 已提交
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
		p += len;
	}
	if (__put_user(0, envp))
		return -EFAULT;
	current->mm->env_end = p;

	/* Put the elf_info on the stack in the right place.  */
	sp = (elf_addr_t __user *)envp + 1;
	if (copy_to_user(sp, elf_info, ei_index * sizeof(elf_addr_t)))
		return -EFAULT;
	return 0;
}

#ifndef elf_map

static unsigned long elf_map(struct file *filep, unsigned long addr,
J
Jiri Kosina 已提交
321 322
		struct elf_phdr *eppnt, int prot, int type,
		unsigned long total_size)
L
Linus Torvalds 已提交
323 324
{
	unsigned long map_addr;
J
Jiri Kosina 已提交
325 326 327 328
	unsigned long size = eppnt->p_filesz + ELF_PAGEOFFSET(eppnt->p_vaddr);
	unsigned long off = eppnt->p_offset - ELF_PAGEOFFSET(eppnt->p_vaddr);
	addr = ELF_PAGESTART(addr);
	size = ELF_PAGEALIGN(size);
L
Linus Torvalds 已提交
329

330 331
	/* mmap() will return -EINVAL if given a zero size, but a
	 * segment with zero filesize is perfectly valid */
J
Jiri Kosina 已提交
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
	if (!size)
		return addr;

	down_write(&current->mm->mmap_sem);
	/*
	* total_size is the size of the ELF (interpreter) image.
	* The _first_ mmap needs to know the full size, otherwise
	* randomization might put this image into an overlapping
	* position with the ELF binary image. (since size < total_size)
	* So we first map the 'big' image - and unmap the remainder at
	* the end. (which unmap is needed for ELF images with holes.)
	*/
	if (total_size) {
		total_size = ELF_PAGEALIGN(total_size);
		map_addr = do_mmap(filep, addr, total_size, prot, type, off);
		if (!BAD_ADDR(map_addr))
			do_munmap(current->mm, map_addr+size, total_size-size);
	} else
		map_addr = do_mmap(filep, addr, size, prot, type, off);

L
Linus Torvalds 已提交
352 353 354 355 356 357
	up_write(&current->mm->mmap_sem);
	return(map_addr);
}

#endif /* !elf_map */

J
Jiri Kosina 已提交
358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376
static unsigned long total_mapping_size(struct elf_phdr *cmds, int nr)
{
	int i, first_idx = -1, last_idx = -1;

	for (i = 0; i < nr; i++) {
		if (cmds[i].p_type == PT_LOAD) {
			last_idx = i;
			if (first_idx == -1)
				first_idx = i;
		}
	}
	if (first_idx == -1)
		return 0;

	return cmds[last_idx].p_vaddr + cmds[last_idx].p_memsz -
				ELF_PAGESTART(cmds[first_idx].p_vaddr);
}


L
Linus Torvalds 已提交
377 378 379 380 381
/* This is much more generalized than the library routine read function,
   so we keep this separate.  Technically the library read function
   is only provided so that we can read a.out libraries that have
   an ELF header */

382
static unsigned long load_elf_interp(struct elfhdr *interp_elf_ex,
J
Jiri Kosina 已提交
383 384
		struct file *interpreter, unsigned long *interp_map_addr,
		unsigned long no_base)
L
Linus Torvalds 已提交
385 386 387 388 389 390 391
{
	struct elf_phdr *elf_phdata;
	struct elf_phdr *eppnt;
	unsigned long load_addr = 0;
	int load_addr_set = 0;
	unsigned long last_bss = 0, elf_bss = 0;
	unsigned long error = ~0UL;
J
Jiri Kosina 已提交
392
	unsigned long total_size;
L
Linus Torvalds 已提交
393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417
	int retval, i, size;

	/* First of all, some simple consistency checks */
	if (interp_elf_ex->e_type != ET_EXEC &&
	    interp_elf_ex->e_type != ET_DYN)
		goto out;
	if (!elf_check_arch(interp_elf_ex))
		goto out;
	if (!interpreter->f_op || !interpreter->f_op->mmap)
		goto out;

	/*
	 * If the size of this structure has changed, then punt, since
	 * we will be doing the wrong thing.
	 */
	if (interp_elf_ex->e_phentsize != sizeof(struct elf_phdr))
		goto out;
	if (interp_elf_ex->e_phnum < 1 ||
		interp_elf_ex->e_phnum > 65536U / sizeof(struct elf_phdr))
		goto out;

	/* Now read in all of the header information */
	size = sizeof(struct elf_phdr) * interp_elf_ex->e_phnum;
	if (size > ELF_MIN_ALIGN)
		goto out;
418
	elf_phdata = kmalloc(size, GFP_KERNEL);
L
Linus Torvalds 已提交
419 420 421
	if (!elf_phdata)
		goto out;

422 423
	retval = kernel_read(interpreter, interp_elf_ex->e_phoff,
			     (char *)elf_phdata,size);
L
Linus Torvalds 已提交
424 425 426 427 428 429 430
	error = -EIO;
	if (retval != size) {
		if (retval < 0)
			error = retval;	
		goto out_close;
	}

J
Jiri Kosina 已提交
431 432 433 434 435 436
	total_size = total_mapping_size(elf_phdata, interp_elf_ex->e_phnum);
	if (!total_size) {
		error = -EINVAL;
		goto out_close;
	}

L
Linus Torvalds 已提交
437
	eppnt = elf_phdata;
438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
	for (i = 0; i < interp_elf_ex->e_phnum; i++, eppnt++) {
		if (eppnt->p_type == PT_LOAD) {
			int elf_type = MAP_PRIVATE | MAP_DENYWRITE;
			int elf_prot = 0;
			unsigned long vaddr = 0;
			unsigned long k, map_addr;

			if (eppnt->p_flags & PF_R)
		    		elf_prot = PROT_READ;
			if (eppnt->p_flags & PF_W)
				elf_prot |= PROT_WRITE;
			if (eppnt->p_flags & PF_X)
				elf_prot |= PROT_EXEC;
			vaddr = eppnt->p_vaddr;
			if (interp_elf_ex->e_type == ET_EXEC || load_addr_set)
				elf_type |= MAP_FIXED;
J
Jiri Kosina 已提交
454 455
			else if (no_base && interp_elf_ex->e_type == ET_DYN)
				load_addr = -vaddr;
456 457

			map_addr = elf_map(interpreter, load_addr + vaddr,
458
					eppnt, elf_prot, elf_type, total_size);
J
Jiri Kosina 已提交
459 460 461
			total_size = 0;
			if (!*interp_map_addr)
				*interp_map_addr = map_addr;
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477
			error = map_addr;
			if (BAD_ADDR(map_addr))
				goto out_close;

			if (!load_addr_set &&
			    interp_elf_ex->e_type == ET_DYN) {
				load_addr = map_addr - ELF_PAGESTART(vaddr);
				load_addr_set = 1;
			}

			/*
			 * Check to see if the section's size will overflow the
			 * allowed task size. Note that p_filesz must always be
			 * <= p_memsize so it's only necessary to check p_memsz.
			 */
			k = load_addr + eppnt->p_vaddr;
478
			if (BAD_ADDR(k) ||
479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501
			    eppnt->p_filesz > eppnt->p_memsz ||
			    eppnt->p_memsz > TASK_SIZE ||
			    TASK_SIZE - eppnt->p_memsz < k) {
				error = -ENOMEM;
				goto out_close;
			}

			/*
			 * Find the end of the file mapping for this phdr, and
			 * keep track of the largest address we see for this.
			 */
			k = load_addr + eppnt->p_vaddr + eppnt->p_filesz;
			if (k > elf_bss)
				elf_bss = k;

			/*
			 * Do the same thing for the memory mapping - between
			 * elf_bss and last_bss is the bss section.
			 */
			k = load_addr + eppnt->p_memsz + eppnt->p_vaddr;
			if (k > last_bss)
				last_bss = k;
		}
L
Linus Torvalds 已提交
502 503
	}

504 505 506 507 508 509 510 511 512 513 514
	if (last_bss > elf_bss) {
		/*
		 * Now fill out the bss section.  First pad the last page up
		 * to the page boundary, and then perform a mmap to make sure
		 * that there are zero-mapped pages up to and including the
		 * last bss page.
		 */
		if (padzero(elf_bss)) {
			error = -EFAULT;
			goto out_close;
		}
L
Linus Torvalds 已提交
515

516 517
		/* What we have mapped so far */
		elf_bss = ELF_PAGESTART(elf_bss + ELF_MIN_ALIGN - 1);
L
Linus Torvalds 已提交
518

519
		/* Map the last of the bss segment */
L
Linus Torvalds 已提交
520 521 522 523 524 525 526
		down_write(&current->mm->mmap_sem);
		error = do_brk(elf_bss, last_bss - elf_bss);
		up_write(&current->mm->mmap_sem);
		if (BAD_ADDR(error))
			goto out_close;
	}

J
Jiri Kosina 已提交
527
	error = load_addr;
L
Linus Torvalds 已提交
528 529 530 531 532 533 534 535 536 537 538 539 540 541 542

out_close:
	kfree(elf_phdata);
out:
	return error;
}

/*
 * These are the functions used to load ELF style executables and shared
 * libraries.  There is no binary dependent code anywhere else.
 */

#define INTERPRETER_NONE 0
#define INTERPRETER_ELF 2

543
#ifndef STACK_RND_MASK
544
#define STACK_RND_MASK (0x7ff >> (PAGE_SHIFT - 12))	/* 8MB of VA */
545
#endif
L
Linus Torvalds 已提交
546 547 548 549 550

static unsigned long randomize_stack_top(unsigned long stack_top)
{
	unsigned int random_variable = 0;

551 552
	if ((current->flags & PF_RANDOMIZE) &&
		!(current->personality & ADDR_NO_RANDOMIZE)) {
553 554 555
		random_variable = get_random_int() & STACK_RND_MASK;
		random_variable <<= PAGE_SHIFT;
	}
L
Linus Torvalds 已提交
556
#ifdef CONFIG_STACK_GROWSUP
557
	return PAGE_ALIGN(stack_top) + random_variable;
L
Linus Torvalds 已提交
558
#else
559
	return PAGE_ALIGN(stack_top) - random_variable;
L
Linus Torvalds 已提交
560 561 562
#endif
}

563
static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs)
L
Linus Torvalds 已提交
564 565 566 567 568 569
{
	struct file *interpreter = NULL; /* to shut gcc up */
 	unsigned long load_addr = 0, load_bias = 0;
	int load_addr_set = 0;
	char * elf_interpreter = NULL;
	unsigned long error;
570
	struct elf_phdr *elf_ppnt, *elf_phdata;
L
Linus Torvalds 已提交
571 572 573
	unsigned long elf_bss, elf_brk;
	int retval, i;
	unsigned int size;
J
Jiri Kosina 已提交
574 575
	unsigned long elf_entry;
	unsigned long interp_load_addr = 0;
L
Linus Torvalds 已提交
576 577
	unsigned long start_code, end_code, start_data, end_data;
	unsigned long reloc_func_desc = 0;
578
	int executable_stack = EXSTACK_DEFAULT;
L
Linus Torvalds 已提交
579 580 581 582 583 584 585 586 587 588 589 590 591
	unsigned long def_flags = 0;
	struct {
		struct elfhdr elf_ex;
		struct elfhdr interp_elf_ex;
	} *loc;

	loc = kmalloc(sizeof(*loc), GFP_KERNEL);
	if (!loc) {
		retval = -ENOMEM;
		goto out_ret;
	}
	
	/* Get the exec-header */
592
	loc->elf_ex = *((struct elfhdr *)bprm->buf);
L
Linus Torvalds 已提交
593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613

	retval = -ENOEXEC;
	/* First of all, some simple consistency checks */
	if (memcmp(loc->elf_ex.e_ident, ELFMAG, SELFMAG) != 0)
		goto out;

	if (loc->elf_ex.e_type != ET_EXEC && loc->elf_ex.e_type != ET_DYN)
		goto out;
	if (!elf_check_arch(&loc->elf_ex))
		goto out;
	if (!bprm->file->f_op||!bprm->file->f_op->mmap)
		goto out;

	/* Now read in all of the header information */
	if (loc->elf_ex.e_phentsize != sizeof(struct elf_phdr))
		goto out;
	if (loc->elf_ex.e_phnum < 1 ||
	 	loc->elf_ex.e_phnum > 65536U / sizeof(struct elf_phdr))
		goto out;
	size = loc->elf_ex.e_phnum * sizeof(struct elf_phdr);
	retval = -ENOMEM;
614
	elf_phdata = kmalloc(size, GFP_KERNEL);
L
Linus Torvalds 已提交
615 616 617
	if (!elf_phdata)
		goto out;

618 619
	retval = kernel_read(bprm->file, loc->elf_ex.e_phoff,
			     (char *)elf_phdata, size);
L
Linus Torvalds 已提交
620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643
	if (retval != size) {
		if (retval >= 0)
			retval = -EIO;
		goto out_free_ph;
	}

	elf_ppnt = elf_phdata;
	elf_bss = 0;
	elf_brk = 0;

	start_code = ~0UL;
	end_code = 0;
	start_data = 0;
	end_data = 0;

	for (i = 0; i < loc->elf_ex.e_phnum; i++) {
		if (elf_ppnt->p_type == PT_INTERP) {
			/* This is the program interpreter used for
			 * shared libraries - for now assume that this
			 * is an a.out format binary
			 */
			retval = -ENOEXEC;
			if (elf_ppnt->p_filesz > PATH_MAX || 
			    elf_ppnt->p_filesz < 2)
644
				goto out_free_ph;
L
Linus Torvalds 已提交
645 646

			retval = -ENOMEM;
647
			elf_interpreter = kmalloc(elf_ppnt->p_filesz,
648
						  GFP_KERNEL);
L
Linus Torvalds 已提交
649
			if (!elf_interpreter)
650
				goto out_free_ph;
L
Linus Torvalds 已提交
651 652

			retval = kernel_read(bprm->file, elf_ppnt->p_offset,
653 654
					     elf_interpreter,
					     elf_ppnt->p_filesz);
L
Linus Torvalds 已提交
655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683
			if (retval != elf_ppnt->p_filesz) {
				if (retval >= 0)
					retval = -EIO;
				goto out_free_interp;
			}
			/* make sure path is NULL terminated */
			retval = -ENOEXEC;
			if (elf_interpreter[elf_ppnt->p_filesz - 1] != '\0')
				goto out_free_interp;

			/*
			 * The early SET_PERSONALITY here is so that the lookup
			 * for the interpreter happens in the namespace of the 
			 * to-be-execed image.  SET_PERSONALITY can select an
			 * alternate root.
			 *
			 * However, SET_PERSONALITY is NOT allowed to switch
			 * this task into the new images's memory mapping
			 * policy - that is, TASK_SIZE must still evaluate to
			 * that which is appropriate to the execing application.
			 * This is because exit_mmap() needs to have TASK_SIZE
			 * evaluate to the size of the old image.
			 *
			 * So if (say) a 64-bit application is execing a 32-bit
			 * application it is the architecture's responsibility
			 * to defer changing the value of TASK_SIZE until the
			 * switch really is going to happen - do this in
			 * flush_thread().	- akpm
			 */
684
			SET_PERSONALITY(loc->elf_ex);
L
Linus Torvalds 已提交
685 686 687 688 689

			interpreter = open_exec(elf_interpreter);
			retval = PTR_ERR(interpreter);
			if (IS_ERR(interpreter))
				goto out_free_interp;
690 691 692 693 694 695 696 697 698

			/*
			 * If the binary is not readable then enforce
			 * mm->dumpable = 0 regardless of the interpreter's
			 * permissions.
			 */
			if (file_permission(interpreter, MAY_READ) < 0)
				bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;

699 700
			retval = kernel_read(interpreter, 0, bprm->buf,
					     BINPRM_BUF_SIZE);
L
Linus Torvalds 已提交
701 702 703 704 705 706 707
			if (retval != BINPRM_BUF_SIZE) {
				if (retval >= 0)
					retval = -EIO;
				goto out_free_dentry;
			}

			/* Get the exec headers */
708
			loc->interp_elf_ex = *((struct elfhdr *)bprm->buf);
L
Linus Torvalds 已提交
709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726
			break;
		}
		elf_ppnt++;
	}

	elf_ppnt = elf_phdata;
	for (i = 0; i < loc->elf_ex.e_phnum; i++, elf_ppnt++)
		if (elf_ppnt->p_type == PT_GNU_STACK) {
			if (elf_ppnt->p_flags & PF_X)
				executable_stack = EXSTACK_ENABLE_X;
			else
				executable_stack = EXSTACK_DISABLE_X;
			break;
		}

	/* Some simple consistency checks for the interpreter */
	if (elf_interpreter) {
		retval = -ELIBBAD;
727 728
		/* Not an ELF interpreter */
		if (memcmp(loc->interp_elf_ex.e_ident, ELFMAG, SELFMAG) != 0)
L
Linus Torvalds 已提交
729 730
			goto out_free_dentry;
		/* Verify the interpreter has a valid arch */
731
		if (!elf_check_arch(&loc->interp_elf_ex))
L
Linus Torvalds 已提交
732 733 734
			goto out_free_dentry;
	} else {
		/* Executables without an interpreter also need a personality  */
735
		SET_PERSONALITY(loc->elf_ex);
L
Linus Torvalds 已提交
736 737 738 739 740 741 742 743 744 745 746 747 748
	}

	/* Flush all traces of the currently running executable */
	retval = flush_old_exec(bprm);
	if (retval)
		goto out_free_dentry;

	/* OK, This is the point of no return */
	current->flags &= ~PF_FORKNOEXEC;
	current->mm->def_flags = def_flags;

	/* Do this immediately, since STACK_TOP as used in setup_arg_pages
	   may depend on the personality.  */
749
	SET_PERSONALITY(loc->elf_ex);
L
Linus Torvalds 已提交
750 751 752
	if (elf_read_implies_exec(loc->elf_ex, executable_stack))
		current->personality |= READ_IMPLIES_EXEC;

753
	if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space)
L
Linus Torvalds 已提交
754 755 756 757 758 759
		current->flags |= PF_RANDOMIZE;
	arch_pick_mmap_layout(current->mm);

	/* Do this so that we can load the interpreter, if need be.  We will
	   change some of these later */
	current->mm->free_area_cache = current->mm->mmap_base;
760
	current->mm->cached_hole_size = 0;
L
Linus Torvalds 已提交
761 762 763 764 765 766 767 768 769
	retval = setup_arg_pages(bprm, randomize_stack_top(STACK_TOP),
				 executable_stack);
	if (retval < 0) {
		send_sig(SIGKILL, current, 0);
		goto out_free_dentry;
	}
	
	current->mm->start_stack = bprm->p;

770
	/* Now we do a little grungy work by mmapping the ELF image into
J
Jiri Kosina 已提交
771
	   the correct location in memory. */
772 773
	for(i = 0, elf_ppnt = elf_phdata;
	    i < loc->elf_ex.e_phnum; i++, elf_ppnt++) {
L
Linus Torvalds 已提交
774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800
		int elf_prot = 0, elf_flags;
		unsigned long k, vaddr;

		if (elf_ppnt->p_type != PT_LOAD)
			continue;

		if (unlikely (elf_brk > elf_bss)) {
			unsigned long nbyte;
	            
			/* There was a PT_LOAD segment with p_memsz > p_filesz
			   before this one. Map anonymous pages, if needed,
			   and clear the area.  */
			retval = set_brk (elf_bss + load_bias,
					  elf_brk + load_bias);
			if (retval) {
				send_sig(SIGKILL, current, 0);
				goto out_free_dentry;
			}
			nbyte = ELF_PAGEOFFSET(elf_bss);
			if (nbyte) {
				nbyte = ELF_MIN_ALIGN - nbyte;
				if (nbyte > elf_brk - elf_bss)
					nbyte = elf_brk - elf_bss;
				if (clear_user((void __user *)elf_bss +
							load_bias, nbyte)) {
					/*
					 * This bss-zeroing can fail if the ELF
801
					 * file specifies odd protections. So
L
Linus Torvalds 已提交
802 803 804 805 806 807
					 * we don't check the return value
					 */
				}
			}
		}

808 809 810 811 812 813
		if (elf_ppnt->p_flags & PF_R)
			elf_prot |= PROT_READ;
		if (elf_ppnt->p_flags & PF_W)
			elf_prot |= PROT_WRITE;
		if (elf_ppnt->p_flags & PF_X)
			elf_prot |= PROT_EXEC;
L
Linus Torvalds 已提交
814

815
		elf_flags = MAP_PRIVATE | MAP_DENYWRITE | MAP_EXECUTABLE;
L
Linus Torvalds 已提交
816 817 818 819 820

		vaddr = elf_ppnt->p_vaddr;
		if (loc->elf_ex.e_type == ET_EXEC || load_addr_set) {
			elf_flags |= MAP_FIXED;
		} else if (loc->elf_ex.e_type == ET_DYN) {
821 822 823 824
			/* Try and get dynamic programs out of the way of the
			 * default mmap base, as well as whatever program they
			 * might try to exec.  This is because the brk will
			 * follow the loader, and is not movable.  */
J
Jiri Kosina 已提交
825 826 827
#ifdef CONFIG_X86
			load_bias = 0;
#else
828
			load_bias = ELF_PAGESTART(ELF_ET_DYN_BASE - vaddr);
J
Jiri Kosina 已提交
829
#endif
L
Linus Torvalds 已提交
830 831
		}

832
		error = elf_map(bprm->file, load_bias + vaddr, elf_ppnt,
833
				elf_prot, elf_flags, 0);
L
Linus Torvalds 已提交
834 835
		if (BAD_ADDR(error)) {
			send_sig(SIGKILL, current, 0);
836 837
			retval = IS_ERR((void *)error) ?
				PTR_ERR((void*)error) : -EINVAL;
L
Linus Torvalds 已提交
838 839 840 841 842 843 844 845 846 847 848 849 850 851
			goto out_free_dentry;
		}

		if (!load_addr_set) {
			load_addr_set = 1;
			load_addr = (elf_ppnt->p_vaddr - elf_ppnt->p_offset);
			if (loc->elf_ex.e_type == ET_DYN) {
				load_bias += error -
				             ELF_PAGESTART(load_bias + vaddr);
				load_addr += load_bias;
				reloc_func_desc = load_bias;
			}
		}
		k = elf_ppnt->p_vaddr;
852 853 854 855
		if (k < start_code)
			start_code = k;
		if (start_data < k)
			start_data = k;
L
Linus Torvalds 已提交
856 857 858 859 860 861

		/*
		 * Check to see if the section's size will overflow the
		 * allowed task size. Note that p_filesz must always be
		 * <= p_memsz so it is only necessary to check p_memsz.
		 */
862
		if (BAD_ADDR(k) || elf_ppnt->p_filesz > elf_ppnt->p_memsz ||
L
Linus Torvalds 已提交
863 864
		    elf_ppnt->p_memsz > TASK_SIZE ||
		    TASK_SIZE - elf_ppnt->p_memsz < k) {
865
			/* set_brk can never work. Avoid overflows. */
L
Linus Torvalds 已提交
866
			send_sig(SIGKILL, current, 0);
867
			retval = -EINVAL;
L
Linus Torvalds 已提交
868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901
			goto out_free_dentry;
		}

		k = elf_ppnt->p_vaddr + elf_ppnt->p_filesz;

		if (k > elf_bss)
			elf_bss = k;
		if ((elf_ppnt->p_flags & PF_X) && end_code < k)
			end_code = k;
		if (end_data < k)
			end_data = k;
		k = elf_ppnt->p_vaddr + elf_ppnt->p_memsz;
		if (k > elf_brk)
			elf_brk = k;
	}

	loc->elf_ex.e_entry += load_bias;
	elf_bss += load_bias;
	elf_brk += load_bias;
	start_code += load_bias;
	end_code += load_bias;
	start_data += load_bias;
	end_data += load_bias;

	/* Calling set_brk effectively mmaps the pages that we need
	 * for the bss and break sections.  We must do this before
	 * mapping in the interpreter, to make sure it doesn't wind
	 * up getting placed where the bss needs to go.
	 */
	retval = set_brk(elf_bss, elf_brk);
	if (retval) {
		send_sig(SIGKILL, current, 0);
		goto out_free_dentry;
	}
A
akpm@osdl.org 已提交
902
	if (likely(elf_bss != elf_brk) && unlikely(padzero(elf_bss))) {
L
Linus Torvalds 已提交
903 904 905 906 907 908
		send_sig(SIGSEGV, current, 0);
		retval = -EFAULT; /* Nobody gets to see this, but.. */
		goto out_free_dentry;
	}

	if (elf_interpreter) {
909 910 911 912 913 914 915 916 917 918 919 920 921
		unsigned long uninitialized_var(interp_map_addr);

		elf_entry = load_elf_interp(&loc->interp_elf_ex,
					    interpreter,
					    &interp_map_addr,
					    load_bias);
		if (!IS_ERR((void *)elf_entry)) {
			/*
			 * load_elf_interp() returns relocation
			 * adjustment
			 */
			interp_load_addr = elf_entry;
			elf_entry += loc->interp_elf_ex.e_entry;
J
Jiri Kosina 已提交
922
		}
L
Linus Torvalds 已提交
923 924
		if (BAD_ADDR(elf_entry)) {
			force_sig(SIGSEGV, current);
925 926
			retval = IS_ERR((void *)elf_entry) ?
					(int)elf_entry : -EINVAL;
L
Linus Torvalds 已提交
927 928 929 930 931 932 933 934 935
			goto out_free_dentry;
		}
		reloc_func_desc = interp_load_addr;

		allow_write_access(interpreter);
		fput(interpreter);
		kfree(elf_interpreter);
	} else {
		elf_entry = loc->elf_ex.e_entry;
936
		if (BAD_ADDR(elf_entry)) {
937 938
			force_sig(SIGSEGV, current);
			retval = -EINVAL;
939 940
			goto out_free_dentry;
		}
L
Linus Torvalds 已提交
941 942 943 944 945 946
	}

	kfree(elf_phdata);

	set_binfmt(&elf_format);

947
#ifdef ARCH_HAS_SETUP_ADDITIONAL_PAGES
948
	retval = arch_setup_additional_pages(bprm, !!elf_interpreter);
949 950
	if (retval < 0) {
		send_sig(SIGKILL, current, 0);
951
		goto out;
952 953 954
	}
#endif /* ARCH_HAS_SETUP_ADDITIONAL_PAGES */

955
	install_exec_creds(bprm);
L
Linus Torvalds 已提交
956
	current->flags &= ~PF_FORKNOEXEC;
957
	retval = create_elf_tables(bprm, &loc->elf_ex,
958
			  load_addr, interp_load_addr);
959 960 961 962
	if (retval < 0) {
		send_sig(SIGKILL, current, 0);
		goto out;
	}
L
Linus Torvalds 已提交
963 964 965 966 967 968 969
	/* N.B. passed_fileno might not be initialized? */
	current->mm->end_code = end_code;
	current->mm->start_code = start_code;
	current->mm->start_data = start_data;
	current->mm->end_data = end_data;
	current->mm->start_stack = bprm->p;

J
Jiri Kosina 已提交
970
#ifdef arch_randomize_brk
971
	if ((current->flags & PF_RANDOMIZE) && (randomize_va_space > 1))
J
Jiri Kosina 已提交
972 973 974 975
		current->mm->brk = current->mm->start_brk =
			arch_randomize_brk(current->mm);
#endif

L
Linus Torvalds 已提交
976 977 978 979
	if (current->personality & MMAP_PAGE_ZERO) {
		/* Why this, you ask???  Well SVr4 maps page 0 as read-only,
		   and some applications "depend" upon this behavior.
		   Since we do not have the power to recompile these, we
980
		   emulate the SVr4 behavior. Sigh. */
L
Linus Torvalds 已提交
981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013
		down_write(&current->mm->mmap_sem);
		error = do_mmap(NULL, 0, PAGE_SIZE, PROT_READ | PROT_EXEC,
				MAP_FIXED | MAP_PRIVATE, 0);
		up_write(&current->mm->mmap_sem);
	}

#ifdef ELF_PLAT_INIT
	/*
	 * The ABI may specify that certain registers be set up in special
	 * ways (on i386 %edx is the address of a DT_FINI function, for
	 * example.  In addition, it may also specify (eg, PowerPC64 ELF)
	 * that the e_entry field is the address of the function descriptor
	 * for the startup routine, rather than the address of the startup
	 * routine itself.  This macro performs whatever initialization to
	 * the regs structure is required as well as any relocations to the
	 * function descriptor entries when executing dynamically links apps.
	 */
	ELF_PLAT_INIT(regs, reloc_func_desc);
#endif

	start_thread(regs, elf_entry, bprm->p);
	retval = 0;
out:
	kfree(loc);
out_ret:
	return retval;

	/* error cleanup */
out_free_dentry:
	allow_write_access(interpreter);
	if (interpreter)
		fput(interpreter);
out_free_interp:
J
Jesper Juhl 已提交
1014
	kfree(elf_interpreter);
L
Linus Torvalds 已提交
1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030
out_free_ph:
	kfree(elf_phdata);
	goto out;
}

/* This is really simpleminded and specialized - we are loading an
   a.out library that is given an ELF header. */
static int load_elf_library(struct file *file)
{
	struct elf_phdr *elf_phdata;
	struct elf_phdr *eppnt;
	unsigned long elf_bss, bss, len;
	int retval, error, i, j;
	struct elfhdr elf_ex;

	error = -ENOEXEC;
1031
	retval = kernel_read(file, 0, (char *)&elf_ex, sizeof(elf_ex));
L
Linus Torvalds 已提交
1032 1033 1034 1035 1036 1037 1038 1039
	if (retval != sizeof(elf_ex))
		goto out;

	if (memcmp(elf_ex.e_ident, ELFMAG, SELFMAG) != 0)
		goto out;

	/* First of all, some simple consistency checks */
	if (elf_ex.e_type != ET_EXEC || elf_ex.e_phnum > 2 ||
1040
	    !elf_check_arch(&elf_ex) || !file->f_op || !file->f_op->mmap)
L
Linus Torvalds 已提交
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087
		goto out;

	/* Now read in all of the header information */

	j = sizeof(struct elf_phdr) * elf_ex.e_phnum;
	/* j < ELF_MIN_ALIGN because elf_ex.e_phnum <= 2 */

	error = -ENOMEM;
	elf_phdata = kmalloc(j, GFP_KERNEL);
	if (!elf_phdata)
		goto out;

	eppnt = elf_phdata;
	error = -ENOEXEC;
	retval = kernel_read(file, elf_ex.e_phoff, (char *)eppnt, j);
	if (retval != j)
		goto out_free_ph;

	for (j = 0, i = 0; i<elf_ex.e_phnum; i++)
		if ((eppnt + i)->p_type == PT_LOAD)
			j++;
	if (j != 1)
		goto out_free_ph;

	while (eppnt->p_type != PT_LOAD)
		eppnt++;

	/* Now use mmap to map the library into memory. */
	down_write(&current->mm->mmap_sem);
	error = do_mmap(file,
			ELF_PAGESTART(eppnt->p_vaddr),
			(eppnt->p_filesz +
			 ELF_PAGEOFFSET(eppnt->p_vaddr)),
			PROT_READ | PROT_WRITE | PROT_EXEC,
			MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE,
			(eppnt->p_offset -
			 ELF_PAGEOFFSET(eppnt->p_vaddr)));
	up_write(&current->mm->mmap_sem);
	if (error != ELF_PAGESTART(eppnt->p_vaddr))
		goto out_free_ph;

	elf_bss = eppnt->p_vaddr + eppnt->p_filesz;
	if (padzero(elf_bss)) {
		error = -EFAULT;
		goto out_free_ph;
	}

1088 1089
	len = ELF_PAGESTART(eppnt->p_filesz + eppnt->p_vaddr +
			    ELF_MIN_ALIGN - 1);
L
Linus Torvalds 已提交
1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
	bss = eppnt->p_memsz + eppnt->p_vaddr;
	if (bss > len) {
		down_write(&current->mm->mmap_sem);
		do_brk(len, bss - len);
		up_write(&current->mm->mmap_sem);
	}
	error = 0;

out_free_ph:
	kfree(elf_phdata);
out:
	return error;
}

/*
 * Note that some platforms still use traditional core dumps and not
 * the ELF core dump.  Each platform can select it as appropriate.
 */
1108
#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE)
L
Linus Torvalds 已提交
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124

/*
 * ELF core dumper
 *
 * Modelled on fs/exec.c:aout_core_dump()
 * Jeremy Fitzhardinge <jeremy@sw.oz.au>
 */
/*
 * These are the only things you should do on a core-file: use only these
 * functions to write out all the necessary info.
 */
static int dump_write(struct file *file, const void *addr, int nr)
{
	return file->f_op->write(file, addr, nr, &file->f_pos) == nr;
}

1125
static int dump_seek(struct file *file, loff_t off)
L
Linus Torvalds 已提交
1126
{
1127
	if (file->f_op->llseek && file->f_op->llseek != no_llseek) {
1128
		if (file->f_op->llseek(file, off, SEEK_CUR) < 0)
L
Linus Torvalds 已提交
1129
			return 0;
1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
	} else {
		char *buf = (char *)get_zeroed_page(GFP_KERNEL);
		if (!buf)
			return 0;
		while (off > 0) {
			unsigned long n = off;
			if (n > PAGE_SIZE)
				n = PAGE_SIZE;
			if (!dump_write(file, buf, n))
				return 0;
			off -= n;
		}
		free_page((unsigned long)buf);
	}
L
Linus Torvalds 已提交
1144 1145 1146 1147
	return 1;
}

/*
R
Roland McGrath 已提交
1148
 * Decide what to dump of a segment, part, all or none.
L
Linus Torvalds 已提交
1149
 */
R
Roland McGrath 已提交
1150 1151
static unsigned long vma_dump_size(struct vm_area_struct *vma,
				   unsigned long mm_flags)
L
Linus Torvalds 已提交
1152
{
1153 1154
#define FILTER(type)	(mm_flags & (1UL << MMF_DUMP_##type))

R
Roland McGrath 已提交
1155 1156
	/* The vma can be set up to tell us the answer directly.  */
	if (vma->vm_flags & VM_ALWAYSDUMP)
R
Roland McGrath 已提交
1157
		goto whole;
R
Roland McGrath 已提交
1158

1159 1160 1161 1162 1163 1164 1165 1166
	/* Hugetlb memory check */
	if (vma->vm_flags & VM_HUGETLB) {
		if ((vma->vm_flags & VM_SHARED) && FILTER(HUGETLB_SHARED))
			goto whole;
		if (!(vma->vm_flags & VM_SHARED) && FILTER(HUGETLB_PRIVATE))
			goto whole;
	}

L
Linus Torvalds 已提交
1167 1168 1169 1170
	/* Do not dump I/O mapped devices or special mappings */
	if (vma->vm_flags & (VM_IO | VM_RESERVED))
		return 0;

1171 1172
	/* By default, dump shared memory if mapped from an anonymous file. */
	if (vma->vm_flags & VM_SHARED) {
R
Roland McGrath 已提交
1173 1174 1175 1176
		if (vma->vm_file->f_path.dentry->d_inode->i_nlink == 0 ?
		    FILTER(ANON_SHARED) : FILTER(MAPPED_SHARED))
			goto whole;
		return 0;
1177
	}
L
Linus Torvalds 已提交
1178

R
Roland McGrath 已提交
1179 1180 1181 1182 1183
	/* Dump segments that have been written to.  */
	if (vma->anon_vma && FILTER(ANON_PRIVATE))
		goto whole;
	if (vma->vm_file == NULL)
		return 0;
L
Linus Torvalds 已提交
1184

R
Roland McGrath 已提交
1185 1186 1187 1188 1189 1190 1191 1192
	if (FILTER(MAPPED_PRIVATE))
		goto whole;

	/*
	 * If this looks like the beginning of a DSO or executable mapping,
	 * check for an ELF header.  If we find one, dump the first page to
	 * aid in determining what was mapped here.
	 */
R
Roland McGrath 已提交
1193 1194
	if (FILTER(ELF_HEADERS) &&
	    vma->vm_pgoff == 0 && (vma->vm_flags & VM_READ)) {
R
Roland McGrath 已提交
1195 1196
		u32 __user *header = (u32 __user *) vma->vm_start;
		u32 word;
R
Roland McGrath 已提交
1197
		mm_segment_t fs = get_fs();
R
Roland McGrath 已提交
1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209
		/*
		 * Doing it this way gets the constant folded by GCC.
		 */
		union {
			u32 cmp;
			char elfmag[SELFMAG];
		} magic;
		BUILD_BUG_ON(SELFMAG != sizeof word);
		magic.elfmag[EI_MAG0] = ELFMAG0;
		magic.elfmag[EI_MAG1] = ELFMAG1;
		magic.elfmag[EI_MAG2] = ELFMAG2;
		magic.elfmag[EI_MAG3] = ELFMAG3;
R
Roland McGrath 已提交
1210 1211 1212 1213 1214 1215 1216 1217 1218
		/*
		 * Switch to the user "segment" for get_user(),
		 * then put back what elf_core_dump() had in place.
		 */
		set_fs(USER_DS);
		if (unlikely(get_user(word, header)))
			word = 0;
		set_fs(fs);
		if (word == magic.cmp)
R
Roland McGrath 已提交
1219 1220 1221 1222 1223 1224 1225 1226 1227
			return PAGE_SIZE;
	}

#undef	FILTER

	return 0;

whole:
	return vma->vm_end - vma->vm_start;
L
Linus Torvalds 已提交
1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249
}

/* An ELF note in memory */
struct memelfnote
{
	const char *name;
	int type;
	unsigned int datasz;
	void *data;
};

static int notesize(struct memelfnote *en)
{
	int sz;

	sz = sizeof(struct elf_note);
	sz += roundup(strlen(en->name) + 1, 4);
	sz += roundup(en->datasz, 4);

	return sz;
}

1250 1251
#define DUMP_WRITE(addr, nr, foffset)	\
	do { if (!dump_write(file, (addr), (nr))) return 0; *foffset += (nr); } while(0)
L
Linus Torvalds 已提交
1252

1253
static int alignfile(struct file *file, loff_t *foffset)
L
Linus Torvalds 已提交
1254
{
1255
	static const char buf[4] = { 0, };
1256 1257 1258
	DUMP_WRITE(buf, roundup(*foffset, 4) - *foffset, foffset);
	return 1;
}
L
Linus Torvalds 已提交
1259

1260 1261 1262 1263
static int writenote(struct memelfnote *men, struct file *file,
			loff_t *foffset)
{
	struct elf_note en;
L
Linus Torvalds 已提交
1264 1265 1266 1267
	en.n_namesz = strlen(men->name) + 1;
	en.n_descsz = men->datasz;
	en.n_type = men->type;

1268 1269 1270 1271 1272 1273 1274
	DUMP_WRITE(&en, sizeof(en), foffset);
	DUMP_WRITE(men->name, en.n_namesz, foffset);
	if (!alignfile(file, foffset))
		return 0;
	DUMP_WRITE(men->data, men->datasz, foffset);
	if (!alignfile(file, foffset))
		return 0;
L
Linus Torvalds 已提交
1275 1276 1277 1278 1279 1280 1281 1282 1283

	return 1;
}
#undef DUMP_WRITE

#define DUMP_WRITE(addr, nr)	\
	if ((size += (nr)) > limit || !dump_write(file, (addr), (nr))) \
		goto end_coredump;

R
Roland McGrath 已提交
1284 1285
static void fill_elf_header(struct elfhdr *elf, int segs,
			    u16 machine, u32 flags, u8 osabi)
L
Linus Torvalds 已提交
1286
{
1287 1288
	memset(elf, 0, sizeof(*elf));

L
Linus Torvalds 已提交
1289 1290 1291 1292 1293 1294 1295
	memcpy(elf->e_ident, ELFMAG, SELFMAG);
	elf->e_ident[EI_CLASS] = ELF_CLASS;
	elf->e_ident[EI_DATA] = ELF_DATA;
	elf->e_ident[EI_VERSION] = EV_CURRENT;
	elf->e_ident[EI_OSABI] = ELF_OSABI;

	elf->e_type = ET_CORE;
R
Roland McGrath 已提交
1296
	elf->e_machine = machine;
L
Linus Torvalds 已提交
1297 1298
	elf->e_version = EV_CURRENT;
	elf->e_phoff = sizeof(struct elfhdr);
R
Roland McGrath 已提交
1299
	elf->e_flags = flags;
L
Linus Torvalds 已提交
1300 1301 1302
	elf->e_ehsize = sizeof(struct elfhdr);
	elf->e_phentsize = sizeof(struct elf_phdr);
	elf->e_phnum = segs;
1303

L
Linus Torvalds 已提交
1304 1305 1306
	return;
}

1307
static void fill_elf_note_phdr(struct elf_phdr *phdr, int sz, loff_t offset)
L
Linus Torvalds 已提交
1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330
{
	phdr->p_type = PT_NOTE;
	phdr->p_offset = offset;
	phdr->p_vaddr = 0;
	phdr->p_paddr = 0;
	phdr->p_filesz = sz;
	phdr->p_memsz = 0;
	phdr->p_flags = 0;
	phdr->p_align = 0;
	return;
}

static void fill_note(struct memelfnote *note, const char *name, int type, 
		unsigned int sz, void *data)
{
	note->name = name;
	note->type = type;
	note->datasz = sz;
	note->data = data;
	return;
}

/*
1331 1332
 * fill up all the fields in prstatus from the given task struct, except
 * registers which need to be filled up separately.
L
Linus Torvalds 已提交
1333 1334
 */
static void fill_prstatus(struct elf_prstatus *prstatus,
1335
		struct task_struct *p, long signr)
L
Linus Torvalds 已提交
1336 1337 1338 1339
{
	prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
	prstatus->pr_sigpend = p->pending.signal.sig[0];
	prstatus->pr_sighold = p->blocked.sig[0];
1340 1341 1342
	rcu_read_lock();
	prstatus->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
	rcu_read_unlock();
1343 1344 1345
	prstatus->pr_pid = task_pid_vnr(p);
	prstatus->pr_pgrp = task_pgrp_vnr(p);
	prstatus->pr_sid = task_session_vnr(p);
L
Linus Torvalds 已提交
1346
	if (thread_group_leader(p)) {
1347 1348
		struct task_cputime cputime;

L
Linus Torvalds 已提交
1349
		/*
1350 1351
		 * This is the record for the group leader.  It shows the
		 * group-wide total, not its individual thread total.
L
Linus Torvalds 已提交
1352
		 */
1353 1354 1355
		thread_group_cputime(p, &cputime);
		cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
		cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
L
Linus Torvalds 已提交
1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
	} else {
		cputime_to_timeval(p->utime, &prstatus->pr_utime);
		cputime_to_timeval(p->stime, &prstatus->pr_stime);
	}
	cputime_to_timeval(p->signal->cutime, &prstatus->pr_cutime);
	cputime_to_timeval(p->signal->cstime, &prstatus->pr_cstime);
}

static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p,
		       struct mm_struct *mm)
{
1367
	const struct cred *cred;
1368
	unsigned int i, len;
L
Linus Torvalds 已提交
1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383
	
	/* first copy the parameters from user space */
	memset(psinfo, 0, sizeof(struct elf_prpsinfo));

	len = mm->arg_end - mm->arg_start;
	if (len >= ELF_PRARGSZ)
		len = ELF_PRARGSZ-1;
	if (copy_from_user(&psinfo->pr_psargs,
		           (const char __user *)mm->arg_start, len))
		return -EFAULT;
	for(i = 0; i < len; i++)
		if (psinfo->pr_psargs[i] == 0)
			psinfo->pr_psargs[i] = ' ';
	psinfo->pr_psargs[len] = 0;

1384 1385 1386
	rcu_read_lock();
	psinfo->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
	rcu_read_unlock();
1387 1388 1389
	psinfo->pr_pid = task_pid_vnr(p);
	psinfo->pr_pgrp = task_pgrp_vnr(p);
	psinfo->pr_sid = task_session_vnr(p);
L
Linus Torvalds 已提交
1390 1391 1392

	i = p->state ? ffz(~p->state) + 1 : 0;
	psinfo->pr_state = i;
1393
	psinfo->pr_sname = (i > 5) ? '.' : "RSDTZW"[i];
L
Linus Torvalds 已提交
1394 1395 1396
	psinfo->pr_zomb = psinfo->pr_sname == 'Z';
	psinfo->pr_nice = task_nice(p);
	psinfo->pr_flag = p->flags;
1397 1398 1399 1400 1401
	rcu_read_lock();
	cred = __task_cred(p);
	SET_UID(psinfo->pr_uid, cred->uid);
	SET_GID(psinfo->pr_gid, cred->gid);
	rcu_read_unlock();
L
Linus Torvalds 已提交
1402 1403 1404 1405 1406
	strncpy(psinfo->pr_fname, p->comm, sizeof(psinfo->pr_fname));
	
	return 0;
}

R
Roland McGrath 已提交
1407 1408 1409 1410 1411 1412 1413 1414 1415 1416
static void fill_auxv_note(struct memelfnote *note, struct mm_struct *mm)
{
	elf_addr_t *auxv = (elf_addr_t *) mm->saved_auxv;
	int i = 0;
	do
		i += 2;
	while (auxv[i - 2] != AT_NULL);
	fill_note(note, "CORE", NT_AUXV, i * sizeof(elf_addr_t), auxv);
}

1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434
#ifdef CORE_DUMP_USE_REGSET
#include <linux/regset.h>

struct elf_thread_core_info {
	struct elf_thread_core_info *next;
	struct task_struct *task;
	struct elf_prstatus prstatus;
	struct memelfnote notes[0];
};

struct elf_note_info {
	struct elf_thread_core_info *thread;
	struct memelfnote psinfo;
	struct memelfnote auxv;
	size_t size;
	int thread_notes;
};

1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446
/*
 * When a regset has a writeback hook, we call it on each thread before
 * dumping user memory.  On register window machines, this makes sure the
 * user memory backing the register data is up to date before we read it.
 */
static void do_thread_regset_writeback(struct task_struct *task,
				       const struct user_regset *regset)
{
	if (regset->writeback)
		regset->writeback(task, regset, 1);
}

1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467
static int fill_thread_core_info(struct elf_thread_core_info *t,
				 const struct user_regset_view *view,
				 long signr, size_t *total)
{
	unsigned int i;

	/*
	 * NT_PRSTATUS is the one special case, because the regset data
	 * goes into the pr_reg field inside the note contents, rather
	 * than being the whole note contents.  We fill the reset in here.
	 * We assume that regset 0 is NT_PRSTATUS.
	 */
	fill_prstatus(&t->prstatus, t->task, signr);
	(void) view->regsets[0].get(t->task, &view->regsets[0],
				    0, sizeof(t->prstatus.pr_reg),
				    &t->prstatus.pr_reg, NULL);

	fill_note(&t->notes[0], "CORE", NT_PRSTATUS,
		  sizeof(t->prstatus), &t->prstatus);
	*total += notesize(&t->notes[0]);

1468 1469
	do_thread_regset_writeback(t->task, &view->regsets[0]);

1470 1471 1472 1473 1474 1475 1476
	/*
	 * Each other regset might generate a note too.  For each regset
	 * that has no core_note_type or is inactive, we leave t->notes[i]
	 * all zero and we'll know to skip writing it later.
	 */
	for (i = 1; i < view->n; ++i) {
		const struct user_regset *regset = &view->regsets[i];
1477
		do_thread_regset_writeback(t->task, regset);
1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514
		if (regset->core_note_type &&
		    (!regset->active || regset->active(t->task, regset))) {
			int ret;
			size_t size = regset->n * regset->size;
			void *data = kmalloc(size, GFP_KERNEL);
			if (unlikely(!data))
				return 0;
			ret = regset->get(t->task, regset,
					  0, size, data, NULL);
			if (unlikely(ret))
				kfree(data);
			else {
				if (regset->core_note_type != NT_PRFPREG)
					fill_note(&t->notes[i], "LINUX",
						  regset->core_note_type,
						  size, data);
				else {
					t->prstatus.pr_fpvalid = 1;
					fill_note(&t->notes[i], "CORE",
						  NT_PRFPREG, size, data);
				}
				*total += notesize(&t->notes[i]);
			}
		}
	}

	return 1;
}

static int fill_note_info(struct elfhdr *elf, int phdrs,
			  struct elf_note_info *info,
			  long signr, struct pt_regs *regs)
{
	struct task_struct *dump_task = current;
	const struct user_regset_view *view = task_user_regset_view(dump_task);
	struct elf_thread_core_info *t;
	struct elf_prpsinfo *psinfo;
1515
	struct core_thread *ct;
1516 1517 1518 1519 1520 1521 1522 1523 1524
	unsigned int i;

	info->size = 0;
	info->thread = NULL;

	psinfo = kmalloc(sizeof(*psinfo), GFP_KERNEL);
	if (psinfo == NULL)
		return 0;

A
Amerigo Wang 已提交
1525 1526
	fill_note(&info->psinfo, "CORE", NT_PRPSINFO, sizeof(*psinfo), psinfo);

1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553
	/*
	 * Figure out how many notes we're going to need for each thread.
	 */
	info->thread_notes = 0;
	for (i = 0; i < view->n; ++i)
		if (view->regsets[i].core_note_type != 0)
			++info->thread_notes;

	/*
	 * Sanity check.  We rely on regset 0 being in NT_PRSTATUS,
	 * since it is our one special case.
	 */
	if (unlikely(info->thread_notes == 0) ||
	    unlikely(view->regsets[0].core_note_type != NT_PRSTATUS)) {
		WARN_ON(1);
		return 0;
	}

	/*
	 * Initialize the ELF file header.
	 */
	fill_elf_header(elf, phdrs,
			view->e_machine, view->e_flags, view->ei_osabi);

	/*
	 * Allocate a structure for each thread.
	 */
1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571
	for (ct = &dump_task->mm->core_state->dumper; ct; ct = ct->next) {
		t = kzalloc(offsetof(struct elf_thread_core_info,
				     notes[info->thread_notes]),
			    GFP_KERNEL);
		if (unlikely(!t))
			return 0;

		t->task = ct->task;
		if (ct->task == dump_task || !info->thread) {
			t->next = info->thread;
			info->thread = t;
		} else {
			/*
			 * Make sure to keep the original task at
			 * the head of the list.
			 */
			t->next = info->thread->next;
			info->thread->next = t;
1572
		}
1573
	}
1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648

	/*
	 * Now fill in each thread's information.
	 */
	for (t = info->thread; t != NULL; t = t->next)
		if (!fill_thread_core_info(t, view, signr, &info->size))
			return 0;

	/*
	 * Fill in the two process-wide notes.
	 */
	fill_psinfo(psinfo, dump_task->group_leader, dump_task->mm);
	info->size += notesize(&info->psinfo);

	fill_auxv_note(&info->auxv, current->mm);
	info->size += notesize(&info->auxv);

	return 1;
}

static size_t get_note_info_size(struct elf_note_info *info)
{
	return info->size;
}

/*
 * Write all the notes for each thread.  When writing the first thread, the
 * process-wide notes are interleaved after the first thread-specific note.
 */
static int write_note_info(struct elf_note_info *info,
			   struct file *file, loff_t *foffset)
{
	bool first = 1;
	struct elf_thread_core_info *t = info->thread;

	do {
		int i;

		if (!writenote(&t->notes[0], file, foffset))
			return 0;

		if (first && !writenote(&info->psinfo, file, foffset))
			return 0;
		if (first && !writenote(&info->auxv, file, foffset))
			return 0;

		for (i = 1; i < info->thread_notes; ++i)
			if (t->notes[i].data &&
			    !writenote(&t->notes[i], file, foffset))
				return 0;

		first = 0;
		t = t->next;
	} while (t);

	return 1;
}

static void free_note_info(struct elf_note_info *info)
{
	struct elf_thread_core_info *threads = info->thread;
	while (threads) {
		unsigned int i;
		struct elf_thread_core_info *t = threads;
		threads = t->next;
		WARN_ON(t->notes[0].data && t->notes[0].data != &t->prstatus);
		for (i = 1; i < info->thread_notes; ++i)
			kfree(t->notes[i].data);
		kfree(t);
	}
	kfree(info->psinfo.data);
}

#else

L
Linus Torvalds 已提交
1649 1650 1651 1652 1653 1654 1655 1656
/* Here is the structure in which status of each thread is captured. */
struct elf_thread_status
{
	struct list_head list;
	struct elf_prstatus prstatus;	/* NT_PRSTATUS */
	elf_fpregset_t fpu;		/* NT_PRFPREG */
	struct task_struct *thread;
#ifdef ELF_CORE_COPY_XFPREGS
1657
	elf_fpxregset_t xfpu;		/* ELF_CORE_XFPREG_TYPE */
L
Linus Torvalds 已提交
1658 1659 1660 1661 1662 1663 1664
#endif
	struct memelfnote notes[3];
	int num_notes;
};

/*
 * In order to add the specific thread information for the elf file format,
1665 1666
 * we need to keep a linked list of every threads pr_status and then create
 * a single section for them in the final core file.
L
Linus Torvalds 已提交
1667 1668 1669 1670 1671 1672 1673 1674 1675 1676
 */
static int elf_dump_thread_status(long signr, struct elf_thread_status *t)
{
	int sz = 0;
	struct task_struct *p = t->thread;
	t->num_notes = 0;

	fill_prstatus(&t->prstatus, p, signr);
	elf_core_copy_task_regs(p, &t->prstatus.pr_reg);	
	
1677 1678
	fill_note(&t->notes[0], "CORE", NT_PRSTATUS, sizeof(t->prstatus),
		  &(t->prstatus));
L
Linus Torvalds 已提交
1679 1680 1681
	t->num_notes++;
	sz += notesize(&t->notes[0]);

1682 1683 1684 1685
	if ((t->prstatus.pr_fpvalid = elf_core_copy_task_fpregs(p, NULL,
								&t->fpu))) {
		fill_note(&t->notes[1], "CORE", NT_PRFPREG, sizeof(t->fpu),
			  &(t->fpu));
L
Linus Torvalds 已提交
1686 1687 1688 1689 1690 1691
		t->num_notes++;
		sz += notesize(&t->notes[1]);
	}

#ifdef ELF_CORE_COPY_XFPREGS
	if (elf_core_copy_task_xfpregs(p, &t->xfpu)) {
1692 1693
		fill_note(&t->notes[2], "LINUX", ELF_CORE_XFPREG_TYPE,
			  sizeof(t->xfpu), &t->xfpu);
L
Linus Torvalds 已提交
1694 1695 1696 1697 1698 1699 1700
		t->num_notes++;
		sz += notesize(&t->notes[2]);
	}
#endif	
	return sz;
}

R
Roland McGrath 已提交
1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713
struct elf_note_info {
	struct memelfnote *notes;
	struct elf_prstatus *prstatus;	/* NT_PRSTATUS */
	struct elf_prpsinfo *psinfo;	/* NT_PRPSINFO */
	struct list_head thread_list;
	elf_fpregset_t *fpu;
#ifdef ELF_CORE_COPY_XFPREGS
	elf_fpxregset_t *xfpu;
#endif
	int thread_status_size;
	int numnote;
};

A
Amerigo Wang 已提交
1714
static int elf_note_info_init(struct elf_note_info *info)
R
Roland McGrath 已提交
1715
{
A
Amerigo Wang 已提交
1716
	memset(info, 0, sizeof(*info));
R
Roland McGrath 已提交
1717 1718
	INIT_LIST_HEAD(&info->thread_list);

A
Amerigo Wang 已提交
1719 1720
	/* Allocate space for six ELF notes */
	info->notes = kmalloc(6 * sizeof(struct memelfnote), GFP_KERNEL);
R
Roland McGrath 已提交
1721 1722 1723 1724
	if (!info->notes)
		return 0;
	info->psinfo = kmalloc(sizeof(*info->psinfo), GFP_KERNEL);
	if (!info->psinfo)
A
Amerigo Wang 已提交
1725
		goto notes_free;
R
Roland McGrath 已提交
1726 1727
	info->prstatus = kmalloc(sizeof(*info->prstatus), GFP_KERNEL);
	if (!info->prstatus)
A
Amerigo Wang 已提交
1728
		goto psinfo_free;
R
Roland McGrath 已提交
1729 1730
	info->fpu = kmalloc(sizeof(*info->fpu), GFP_KERNEL);
	if (!info->fpu)
A
Amerigo Wang 已提交
1731
		goto prstatus_free;
R
Roland McGrath 已提交
1732 1733 1734
#ifdef ELF_CORE_COPY_XFPREGS
	info->xfpu = kmalloc(sizeof(*info->xfpu), GFP_KERNEL);
	if (!info->xfpu)
A
Amerigo Wang 已提交
1735
		goto fpu_free;
R
Roland McGrath 已提交
1736
#endif
A
Amerigo Wang 已提交
1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758
	return 1;
#ifdef ELF_CORE_COPY_XFPREGS
 fpu_free:
	kfree(info->fpu);
#endif
 prstatus_free:
	kfree(info->prstatus);
 psinfo_free:
	kfree(info->psinfo);
 notes_free:
	kfree(info->notes);
	return 0;
}

static int fill_note_info(struct elfhdr *elf, int phdrs,
			  struct elf_note_info *info,
			  long signr, struct pt_regs *regs)
{
	struct list_head *t;

	if (!elf_note_info_init(info))
		return 0;
R
Roland McGrath 已提交
1759 1760

	if (signr) {
1761
		struct core_thread *ct;
1762
		struct elf_thread_status *ets;
1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773

		for (ct = current->mm->core_state->dumper.next;
						ct; ct = ct->next) {
			ets = kzalloc(sizeof(*ets), GFP_KERNEL);
			if (!ets)
				return 0;

			ets->thread = ct->task;
			list_add(&ets->list, &info->thread_list);
		}

R
Roland McGrath 已提交
1774 1775 1776
		list_for_each(t, &info->thread_list) {
			int sz;

1777 1778
			ets = list_entry(t, struct elf_thread_status, list);
			sz = elf_dump_thread_status(signr, ets);
R
Roland McGrath 已提交
1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873
			info->thread_status_size += sz;
		}
	}
	/* now collect the dump for the current */
	memset(info->prstatus, 0, sizeof(*info->prstatus));
	fill_prstatus(info->prstatus, current, signr);
	elf_core_copy_regs(&info->prstatus->pr_reg, regs);

	/* Set up header */
	fill_elf_header(elf, phdrs, ELF_ARCH, ELF_CORE_EFLAGS, ELF_OSABI);

	/*
	 * Set up the notes in similar form to SVR4 core dumps made
	 * with info from their /proc.
	 */

	fill_note(info->notes + 0, "CORE", NT_PRSTATUS,
		  sizeof(*info->prstatus), info->prstatus);
	fill_psinfo(info->psinfo, current->group_leader, current->mm);
	fill_note(info->notes + 1, "CORE", NT_PRPSINFO,
		  sizeof(*info->psinfo), info->psinfo);

	info->numnote = 2;

	fill_auxv_note(&info->notes[info->numnote++], current->mm);

	/* Try to dump the FPU. */
	info->prstatus->pr_fpvalid = elf_core_copy_task_fpregs(current, regs,
							       info->fpu);
	if (info->prstatus->pr_fpvalid)
		fill_note(info->notes + info->numnote++,
			  "CORE", NT_PRFPREG, sizeof(*info->fpu), info->fpu);
#ifdef ELF_CORE_COPY_XFPREGS
	if (elf_core_copy_task_xfpregs(current, info->xfpu))
		fill_note(info->notes + info->numnote++,
			  "LINUX", ELF_CORE_XFPREG_TYPE,
			  sizeof(*info->xfpu), info->xfpu);
#endif

	return 1;
}

static size_t get_note_info_size(struct elf_note_info *info)
{
	int sz = 0;
	int i;

	for (i = 0; i < info->numnote; i++)
		sz += notesize(info->notes + i);

	sz += info->thread_status_size;

	return sz;
}

static int write_note_info(struct elf_note_info *info,
			   struct file *file, loff_t *foffset)
{
	int i;
	struct list_head *t;

	for (i = 0; i < info->numnote; i++)
		if (!writenote(info->notes + i, file, foffset))
			return 0;

	/* write out the thread status notes section */
	list_for_each(t, &info->thread_list) {
		struct elf_thread_status *tmp =
				list_entry(t, struct elf_thread_status, list);

		for (i = 0; i < tmp->num_notes; i++)
			if (!writenote(&tmp->notes[i], file, foffset))
				return 0;
	}

	return 1;
}

static void free_note_info(struct elf_note_info *info)
{
	while (!list_empty(&info->thread_list)) {
		struct list_head *tmp = info->thread_list.next;
		list_del(tmp);
		kfree(list_entry(tmp, struct elf_thread_status, list));
	}

	kfree(info->prstatus);
	kfree(info->psinfo);
	kfree(info->notes);
	kfree(info->fpu);
#ifdef ELF_CORE_COPY_XFPREGS
	kfree(info->xfpu);
#endif
}

1874 1875
#endif

1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901
static struct vm_area_struct *first_vma(struct task_struct *tsk,
					struct vm_area_struct *gate_vma)
{
	struct vm_area_struct *ret = tsk->mm->mmap;

	if (ret)
		return ret;
	return gate_vma;
}
/*
 * Helper function for iterating across a vma list.  It ensures that the caller
 * will visit `gate_vma' prior to terminating the search.
 */
static struct vm_area_struct *next_vma(struct vm_area_struct *this_vma,
					struct vm_area_struct *gate_vma)
{
	struct vm_area_struct *ret;

	ret = this_vma->vm_next;
	if (ret)
		return ret;
	if (this_vma == gate_vma)
		return NULL;
	return gate_vma;
}

L
Linus Torvalds 已提交
1902 1903 1904 1905 1906 1907 1908
/*
 * Actual dumper
 *
 * This is a two-pass process; first we find the offsets of the bits,
 * and then they are actually written out.  If we run out of core limit
 * we just truncate.
 */
1909
static int elf_core_dump(long signr, struct pt_regs *regs, struct file *file, unsigned long limit)
L
Linus Torvalds 已提交
1910 1911 1912 1913 1914
{
	int has_dumped = 0;
	mm_segment_t fs;
	int segs;
	size_t size = 0;
1915
	struct vm_area_struct *vma, *gate_vma;
L
Linus Torvalds 已提交
1916
	struct elfhdr *elf = NULL;
1917
	loff_t offset = 0, dataoff, foffset;
1918
	unsigned long mm_flags;
R
Roland McGrath 已提交
1919
	struct elf_note_info info;
L
Linus Torvalds 已提交
1920 1921 1922 1923

	/*
	 * We no longer stop all VM operations.
	 * 
1924 1925 1926
	 * This is because those proceses that could possibly change map_count
	 * or the mmap / vma pages are now blocked in do_exit on current
	 * finishing this core dump.
L
Linus Torvalds 已提交
1927 1928
	 *
	 * Only ptrace can touch these memory addresses, but it doesn't change
1929
	 * the map_count or the pages allocated. So no possibility of crashing
L
Linus Torvalds 已提交
1930 1931 1932 1933 1934 1935
	 * exists while dumping the mm->vm_next areas to the core file.
	 */
  
	/* alloc memory for large data structures: too large to be on stack */
	elf = kmalloc(sizeof(*elf), GFP_KERNEL);
	if (!elf)
1936
		goto out;
1937 1938 1939 1940
	/*
	 * The number of segs are recored into ELF header as 16bit value.
	 * Please check DEFAULT_MAX_MAP_COUNT definition when you modify here.
	 */
L
Linus Torvalds 已提交
1941 1942 1943 1944 1945
	segs = current->mm->map_count;
#ifdef ELF_CORE_EXTRA_PHDRS
	segs += ELF_CORE_EXTRA_PHDRS;
#endif

1946 1947 1948 1949
	gate_vma = get_gate_vma(current);
	if (gate_vma != NULL)
		segs++;

L
Linus Torvalds 已提交
1950
	/*
R
Roland McGrath 已提交
1951 1952
	 * Collect all the non-memory information about the process for the
	 * notes.  This also sets up the file header.
L
Linus Torvalds 已提交
1953
	 */
R
Roland McGrath 已提交
1954 1955 1956
	if (!fill_note_info(elf, segs + 1, /* including notes section */
			    &info, signr, regs))
		goto cleanup;
L
Linus Torvalds 已提交
1957

R
Roland McGrath 已提交
1958 1959
	has_dumped = 1;
	current->flags |= PF_DUMPCORE;
L
Linus Torvalds 已提交
1960 1961 1962 1963 1964 1965
  
	fs = get_fs();
	set_fs(KERNEL_DS);

	DUMP_WRITE(elf, sizeof(*elf));
	offset += sizeof(*elf);				/* Elf header */
1966 1967
	offset += (segs + 1) * sizeof(struct elf_phdr); /* Program headers */
	foffset = offset;
L
Linus Torvalds 已提交
1968 1969 1970 1971

	/* Write notes phdr entry */
	{
		struct elf_phdr phdr;
R
Roland McGrath 已提交
1972
		size_t sz = get_note_info_size(&info);
L
Linus Torvalds 已提交
1973

1974
		sz += elf_coredump_extra_notes_size();
1975

L
Linus Torvalds 已提交
1976 1977 1978 1979 1980 1981 1982
		fill_elf_note_phdr(&phdr, sz, offset);
		offset += sz;
		DUMP_WRITE(&phdr, sizeof(phdr));
	}

	dataoff = offset = roundup(offset, ELF_EXEC_PAGESIZE);

1983 1984 1985 1986 1987 1988 1989
	/*
	 * We must use the same mm->flags while dumping core to avoid
	 * inconsistency between the program headers and bodies, otherwise an
	 * unusable core file can be generated.
	 */
	mm_flags = current->mm->flags;

L
Linus Torvalds 已提交
1990
	/* Write program headers for segments dump */
1991 1992
	for (vma = first_vma(current, gate_vma); vma != NULL;
			vma = next_vma(vma, gate_vma)) {
L
Linus Torvalds 已提交
1993 1994 1995 1996 1997 1998
		struct elf_phdr phdr;

		phdr.p_type = PT_LOAD;
		phdr.p_offset = offset;
		phdr.p_vaddr = vma->vm_start;
		phdr.p_paddr = 0;
R
Roland McGrath 已提交
1999 2000
		phdr.p_filesz = vma_dump_size(vma, mm_flags);
		phdr.p_memsz = vma->vm_end - vma->vm_start;
L
Linus Torvalds 已提交
2001 2002
		offset += phdr.p_filesz;
		phdr.p_flags = vma->vm_flags & VM_READ ? PF_R : 0;
2003 2004 2005 2006
		if (vma->vm_flags & VM_WRITE)
			phdr.p_flags |= PF_W;
		if (vma->vm_flags & VM_EXEC)
			phdr.p_flags |= PF_X;
L
Linus Torvalds 已提交
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
		phdr.p_align = ELF_EXEC_PAGESIZE;

		DUMP_WRITE(&phdr, sizeof(phdr));
	}

#ifdef ELF_CORE_WRITE_EXTRA_PHDRS
	ELF_CORE_WRITE_EXTRA_PHDRS;
#endif

 	/* write out the notes section */
R
Roland McGrath 已提交
2017 2018
	if (!write_note_info(&info, file, &foffset))
		goto end_coredump;
L
Linus Torvalds 已提交
2019

2020 2021
	if (elf_coredump_extra_notes_write(file, &foffset))
		goto end_coredump;
2022

2023
	/* Align to page */
H
Hugh Dickins 已提交
2024 2025
	if (!dump_seek(file, dataoff - foffset))
		goto end_coredump;
L
Linus Torvalds 已提交
2026

2027 2028
	for (vma = first_vma(current, gate_vma); vma != NULL;
			vma = next_vma(vma, gate_vma)) {
L
Linus Torvalds 已提交
2029
		unsigned long addr;
R
Roland McGrath 已提交
2030
		unsigned long end;
L
Linus Torvalds 已提交
2031

R
Roland McGrath 已提交
2032
		end = vma->vm_start + vma_dump_size(vma, mm_flags);
L
Linus Torvalds 已提交
2033

R
Roland McGrath 已提交
2034
		for (addr = vma->vm_start; addr < end; addr += PAGE_SIZE) {
2035
			struct page *page;
H
Hugh Dickins 已提交
2036 2037 2038 2039 2040 2041 2042 2043
			int stop;

			page = get_dump_page(addr);
			if (page) {
				void *kaddr = kmap(page);
				stop = ((size += PAGE_SIZE) > limit) ||
					!dump_write(file, kaddr, PAGE_SIZE);
				kunmap(page);
L
Linus Torvalds 已提交
2044
				page_cache_release(page);
H
Hugh Dickins 已提交
2045 2046 2047 2048
			} else
				stop = !dump_seek(file, PAGE_SIZE);
			if (stop)
				goto end_coredump;
L
Linus Torvalds 已提交
2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059
		}
	}

#ifdef ELF_CORE_WRITE_EXTRA_DATA
	ELF_CORE_WRITE_EXTRA_DATA;
#endif

end_coredump:
	set_fs(fs);

cleanup:
R
Roland McGrath 已提交
2060
	free_note_info(&info);
2061 2062
	kfree(elf);
out:
L
Linus Torvalds 已提交
2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081
	return has_dumped;
}

#endif		/* USE_ELF_CORE_DUMP */

static int __init init_elf_binfmt(void)
{
	return register_binfmt(&elf_format);
}

static void __exit exit_elf_binfmt(void)
{
	/* Remove the COFF and ELF loaders. */
	unregister_binfmt(&elf_format);
}

core_initcall(init_elf_binfmt);
module_exit(exit_elf_binfmt);
MODULE_LICENSE("GPL");