util.c 13.2 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7 8 9 10
/*
 * linux/ipc/util.c
 * Copyright (C) 1992 Krishna Balasubramanian
 *
 * Sep 1997 - Call suser() last after "normal" permission checks so we
 *            get BSD style process accounting right.
 *            Occurs in several places in the IPC code.
 *            Chris Evans, <chris@ferret.lmh.ox.ac.uk>
 * Nov 1999 - ipc helper functions, unified SMP locking
 *	      Manfred Spraul <manfreds@colorfullife.com>
11 12
 * Oct 2002 - One lock per IPC id. RCU ipc_free for lock-free grow_ary().
 *            Mingming Cao <cmm@us.ibm.com>
Linus Torvalds's avatar
Linus Torvalds committed
13 14 15 16 17 18 19 20 21
 */

#include <linux/config.h>
#include <linux/mm.h>
#include <linux/shm.h>
#include <linux/init.h>
#include <linux/msg.h>
#include <linux/smp_lock.h>
#include <linux/vmalloc.h>
Linus Torvalds's avatar
Linus Torvalds committed
22
#include <linux/slab.h>
Linus Torvalds's avatar
Linus Torvalds committed
23
#include <linux/highuid.h>
24
#include <linux/security.h>
Andrew Morton's avatar
Andrew Morton committed
25
#include <linux/rcupdate.h>
26
#include <linux/workqueue.h>
Linus Torvalds's avatar
Linus Torvalds committed
27

28 29
#include <asm/unistd.h>

Linus Torvalds's avatar
Linus Torvalds committed
30 31 32 33 34 35 36 37 38
#include "util.h"

/**
 *	ipc_init	-	initialise IPC subsystem
 *
 *	The various system5 IPC resources (semaphores, messages and shared
 *	memory are initialised
 */
 
39
static int __init ipc_init(void)
Linus Torvalds's avatar
Linus Torvalds committed
40 41 42 43
{
	sem_init();
	msg_init();
	shm_init();
44
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
45
}
46
__initcall(ipc_init);
Linus Torvalds's avatar
Linus Torvalds committed
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

/**
 *	ipc_init_ids		-	initialise IPC identifiers
 *	@ids: Identifier set
 *	@size: Number of identifiers
 *
 *	Given a size for the ipc identifier range (limited below IPCMNI)
 *	set up the sequence range to use then allocate and initialise the
 *	array itself. 
 */
 
void __init ipc_init_ids(struct ipc_ids* ids, int size)
{
	int i;
	sema_init(&ids->sem,1);

	if(size > IPCMNI)
		size = IPCMNI;
	ids->in_use = 0;
	ids->max_id = -1;
	ids->seq = 0;
	{
		int seq_limit = INT_MAX/SEQ_MULTIPLIER;
		if(seq_limit > USHRT_MAX)
			ids->seq_max = USHRT_MAX;
		 else
		 	ids->seq_max = seq_limit;
	}

76 77
	ids->entries = ipc_rcu_alloc(sizeof(struct kern_ipc_perm *)*size +
				     sizeof(struct ipc_id_ary));
Linus Torvalds's avatar
Linus Torvalds committed
78 79 80

	if(ids->entries == NULL) {
		printk(KERN_ERR "ipc_init_ids() failed, ipc service disabled.\n");
81 82
		size = 0;
		ids->entries = &ids->nullentry;
Linus Torvalds's avatar
Linus Torvalds committed
83
	}
84 85 86
	ids->entries->size = size;
	for(i=0;i<size;i++)
		ids->entries->p[i] = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
87 88 89 90 91 92
}

/**
 *	ipc_findkey	-	find a key in an ipc identifier set	
 *	@ids: Identifier set
 *	@key: The key to find
93 94
 *	
 *	Requires ipc_ids.sem locked.
Linus Torvalds's avatar
Linus Torvalds committed
95 96 97 98 99 100 101
 *	Returns the identifier if found or -1 if not.
 */
 
int ipc_findkey(struct ipc_ids* ids, key_t key)
{
	int id;
	struct kern_ipc_perm* p;
102
	int max_id = ids->max_id;
Linus Torvalds's avatar
Linus Torvalds committed
103

104
	/*
105
	 * rcu_dereference() is not needed here
106 107
	 * since ipc_ids.sem is held
	 */
108
	for (id = 0; id <= max_id; id++) {
109
		p = ids->entries->p[id];
Linus Torvalds's avatar
Linus Torvalds committed
110 111 112 113 114 115 116 117
		if(p==NULL)
			continue;
		if (key == p->key)
			return id;
	}
	return -1;
}

118 119 120
/*
 * Requires ipc_ids.sem locked
 */
Linus Torvalds's avatar
Linus Torvalds committed
121 122
static int grow_ary(struct ipc_ids* ids, int newsize)
{
123 124
	struct ipc_id_ary* new;
	struct ipc_id_ary* old;
Linus Torvalds's avatar
Linus Torvalds committed
125
	int i;
126
	int size = ids->entries->size;
Linus Torvalds's avatar
Linus Torvalds committed
127 128 129

	if(newsize > IPCMNI)
		newsize = IPCMNI;
130
	if(newsize <= size)
Linus Torvalds's avatar
Linus Torvalds committed
131 132
		return newsize;

133 134
	new = ipc_rcu_alloc(sizeof(struct kern_ipc_perm *)*newsize +
			    sizeof(struct ipc_id_ary));
Linus Torvalds's avatar
Linus Torvalds committed
135
	if(new == NULL)
136 137 138 139 140 141
		return size;
	new->size = newsize;
	memcpy(new->p, ids->entries->p, sizeof(struct kern_ipc_perm *)*size +
					sizeof(struct ipc_id_ary));
	for(i=size;i<newsize;i++) {
		new->p[i] = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
142 143
	}
	old = ids->entries;
144 145

	/*
146 147
	 * Use rcu_assign_pointer() to make sure the memcpyed contents
	 * of the new array are visible before the new array becomes visible.
148
	 */
149
	rcu_assign_pointer(ids->entries, new);
150

151
	ipc_rcu_putref(old);
152
	return newsize;
Linus Torvalds's avatar
Linus Torvalds committed
153 154 155 156 157 158 159 160 161 162 163 164
}

/**
 *	ipc_addid 	-	add an IPC identifier
 *	@ids: IPC identifier set
 *	@new: new IPC permission set
 *	@size: new size limit for the id array
 *
 *	Add an entry 'new' to the IPC arrays. The permissions object is
 *	initialised and the first free entry is set up and the id assigned
 *	is returned. The list is returned in a locked state on success.
 *	On failure the list is not locked and -1 is returned.
165 166
 *
 *	Called with ipc_ids.sem held.
Linus Torvalds's avatar
Linus Torvalds committed
167 168 169 170 171 172 173
 */
 
int ipc_addid(struct ipc_ids* ids, struct kern_ipc_perm* new, int size)
{
	int id;

	size = grow_ary(ids,size);
174 175

	/*
176
	 * rcu_dereference()() is not needed here since
177 178
	 * ipc_ids.sem is held
	 */
Linus Torvalds's avatar
Linus Torvalds committed
179
	for (id = 0; id < size; id++) {
180
		if(ids->entries->p[id] == NULL)
Linus Torvalds's avatar
Linus Torvalds committed
181 182 183 184 185 186 187 188 189 190 191 192 193 194 195
			goto found;
	}
	return -1;
found:
	ids->in_use++;
	if (id > ids->max_id)
		ids->max_id = id;

	new->cuid = new->uid = current->euid;
	new->gid = new->cgid = current->egid;

	new->seq = ids->seq++;
	if(ids->seq > ids->seq_max)
		ids->seq = 0;

196
	spin_lock_init(&new->lock);
197 198 199
	new->deleted = 0;
	rcu_read_lock();
	spin_lock(&new->lock);
200
	ids->entries->p[id] = new;
Linus Torvalds's avatar
Linus Torvalds committed
201 202 203 204 205 206 207 208 209 210 211 212
	return id;
}

/**
 *	ipc_rmid	-	remove an IPC identifier
 *	@ids: identifier set
 *	@id: Identifier to remove
 *
 *	The identifier must be valid, and in use. The kernel will panic if
 *	fed an invalid identifier. The entry is removed and internal
 *	variables recomputed. The object associated with the identifier
 *	is returned.
213 214
 *	ipc_ids.sem and the spinlock for this ID is hold before this function
 *	is called, and remain locked on the exit.
Linus Torvalds's avatar
Linus Torvalds committed
215 216 217 218 219 220
 */
 
struct kern_ipc_perm* ipc_rmid(struct ipc_ids* ids, int id)
{
	struct kern_ipc_perm* p;
	int lid = id % SEQ_MULTIPLIER;
221
	if(lid >= ids->entries->size)
Linus Torvalds's avatar
Linus Torvalds committed
222
		BUG();
223 224

	/* 
225
	 * do not need a rcu_dereference()() here to force ordering
226 227
	 * on Alpha, since the ipc_ids.sem is held.
	 */	
228 229
	p = ids->entries->p[lid];
	ids->entries->p[lid] = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
230 231 232 233 234 235 236 237 238
	if(p==NULL)
		BUG();
	ids->in_use--;

	if (lid == ids->max_id) {
		do {
			lid--;
			if(lid == -1)
				break;
239
		} while (ids->entries->p[lid] == NULL);
Linus Torvalds's avatar
Linus Torvalds committed
240 241
		ids->max_id = lid;
	}
242
	p->deleted = 1;
Linus Torvalds's avatar
Linus Torvalds committed
243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
	return p;
}

/**
 *	ipc_alloc	-	allocate ipc space
 *	@size: size desired
 *
 *	Allocate memory from the appropriate pools and return a pointer to it.
 *	NULL is returned if the allocation fails
 */
 
void* ipc_alloc(int size)
{
	void* out;
	if(size > PAGE_SIZE)
		out = vmalloc(size);
	else
		out = kmalloc(size, GFP_KERNEL);
	return out;
}

/**
265
 *	ipc_free        -       free ipc space
Linus Torvalds's avatar
Linus Torvalds committed
266 267 268 269 270 271
 *	@ptr: pointer returned by ipc_alloc
 *	@size: size of block
 *
 *	Free a block created with ipc_alloc. The caller must know the size
 *	used in the allocation call.
 */
272

Linus Torvalds's avatar
Linus Torvalds committed
273 274 275 276 277 278 279 280
void ipc_free(void* ptr, int size)
{
	if(size > PAGE_SIZE)
		vfree(ptr);
	else
		kfree(ptr);
}

281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299
/*
 * rcu allocations:
 * There are three headers that are prepended to the actual allocation:
 * - during use: ipc_rcu_hdr.
 * - during the rcu grace period: ipc_rcu_grace.
 * - [only if vmalloc]: ipc_rcu_sched.
 * Their lifetime doesn't overlap, thus the headers share the same memory.
 * Unlike a normal union, they are right-aligned, thus some container_of
 * forward/backward casting is necessary:
 */
struct ipc_rcu_hdr
{
	int refcount;
	int is_vmalloc;
	void *data[0];
};


struct ipc_rcu_grace
300 301 302 303 304 305
{
	struct rcu_head rcu;
	/* "void *" makes sure alignment of following data is sane. */
	void *data[0];
};

306
struct ipc_rcu_sched
307 308 309 310 311 312
{
	struct work_struct work;
	/* "void *" makes sure alignment of following data is sane. */
	void *data[0];
};

313 314 315 316 317
#define HDRLEN_KMALLOC		(sizeof(struct ipc_rcu_grace) > sizeof(struct ipc_rcu_hdr) ? \
					sizeof(struct ipc_rcu_grace) : sizeof(struct ipc_rcu_hdr))
#define HDRLEN_VMALLOC		(sizeof(struct ipc_rcu_sched) > HDRLEN_KMALLOC ? \
					sizeof(struct ipc_rcu_sched) : HDRLEN_KMALLOC)

318 319 320
static inline int rcu_use_vmalloc(int size)
{
	/* Too big for a single page? */
321
	if (HDRLEN_KMALLOC + size > PAGE_SIZE)
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342
		return 1;
	return 0;
}

/**
 *	ipc_rcu_alloc	-	allocate ipc and rcu space 
 *	@size: size desired
 *
 *	Allocate memory for the rcu header structure +  the object.
 *	Returns the pointer to the object.
 *	NULL is returned if the allocation fails. 
 */
 
void* ipc_rcu_alloc(int size)
{
	void* out;
	/* 
	 * We prepend the allocation with the rcu struct, and
	 * workqueue if necessary (for vmalloc). 
	 */
	if (rcu_use_vmalloc(size)) {
343 344 345 346 347 348
		out = vmalloc(HDRLEN_VMALLOC + size);
		if (out) {
			out += HDRLEN_VMALLOC;
			container_of(out, struct ipc_rcu_hdr, data)->is_vmalloc = 1;
			container_of(out, struct ipc_rcu_hdr, data)->refcount = 1;
		}
349
	} else {
350 351 352 353 354 355
		out = kmalloc(HDRLEN_KMALLOC + size, GFP_KERNEL);
		if (out) {
			out += HDRLEN_KMALLOC;
			container_of(out, struct ipc_rcu_hdr, data)->is_vmalloc = 0;
			container_of(out, struct ipc_rcu_hdr, data)->refcount = 1;
		}
356 357 358 359 360
	}

	return out;
}

361 362 363 364 365
void ipc_rcu_getref(void *ptr)
{
	container_of(ptr, struct ipc_rcu_hdr, data)->refcount++;
}

366 367 368 369 370 371
/**
 *	ipc_schedule_free	- free ipc + rcu space
 * 
 * Since RCU callback function is called in bh,
 * we need to defer the vfree to schedule_work
 */
372
static void ipc_schedule_free(struct rcu_head *head)
373
{
374 375 376 377
	struct ipc_rcu_grace *grace =
		container_of(head, struct ipc_rcu_grace, rcu);
	struct ipc_rcu_sched *sched =
			container_of(&(grace->data[0]), struct ipc_rcu_sched, data[0]);
378

379 380
	INIT_WORK(&sched->work, vfree, sched);
	schedule_work(&sched->work);
381 382
}

383 384 385 386 387 388 389 390
/**
 *	ipc_immediate_free	- free ipc + rcu space
 *
 *	Free from the RCU callback context
 *
 */
static void ipc_immediate_free(struct rcu_head *head)
{
391 392
	struct ipc_rcu_grace *free =
		container_of(head, struct ipc_rcu_grace, rcu);
393 394 395
	kfree(free);
}

396
void ipc_rcu_putref(void *ptr)
397
{
398 399 400 401 402 403
	if (--container_of(ptr, struct ipc_rcu_hdr, data)->refcount > 0)
		return;

	if (container_of(ptr, struct ipc_rcu_hdr, data)->is_vmalloc) {
		call_rcu(&container_of(ptr, struct ipc_rcu_grace, data)->rcu,
				ipc_schedule_free);
404
	} else {
405 406
		call_rcu(&container_of(ptr, struct ipc_rcu_grace, data)->rcu,
				ipc_immediate_free);
407 408 409
	}
}

Linus Torvalds's avatar
Linus Torvalds committed
410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433
/**
 *	ipcperms	-	check IPC permissions
 *	@ipcp: IPC permission set
 *	@flag: desired permission set.
 *
 *	Check user, group, other permissions for access
 *	to ipc resources. return 0 if allowed
 */
 
int ipcperms (struct kern_ipc_perm *ipcp, short flag)
{	/* flag will most probably be 0 or S_...UGO from <linux/stat.h> */
	int requested_mode, granted_mode;

	requested_mode = (flag >> 6) | (flag >> 3) | flag;
	granted_mode = ipcp->mode;
	if (current->euid == ipcp->cuid || current->euid == ipcp->uid)
		granted_mode >>= 6;
	else if (in_group_p(ipcp->cgid) || in_group_p(ipcp->gid))
		granted_mode >>= 3;
	/* is there some bit set in requested_mode but not in granted_mode? */
	if ((requested_mode & ~granted_mode & 0007) && 
	    !capable(CAP_IPC_OWNER))
		return -1;

434
	return security_ipc_permission(ipcp, flag);
Linus Torvalds's avatar
Linus Torvalds committed
435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474
}

/*
 * Functions to convert between the kern_ipc_perm structure and the
 * old/new ipc_perm structures
 */

/**
 *	kernel_to_ipc64_perm	-	convert kernel ipc permissions to user
 *	@in: kernel permissions
 *	@out: new style IPC permissions
 *
 *	Turn the kernel object 'in' into a set of permissions descriptions
 *	for returning to userspace (out).
 */
 

void kernel_to_ipc64_perm (struct kern_ipc_perm *in, struct ipc64_perm *out)
{
	out->key	= in->key;
	out->uid	= in->uid;
	out->gid	= in->gid;
	out->cuid	= in->cuid;
	out->cgid	= in->cgid;
	out->mode	= in->mode;
	out->seq	= in->seq;
}

/**
 *	ipc64_perm_to_ipc_perm	-	convert old ipc permissions to new
 *	@in: new style IPC permissions
 *	@out: old style IPC permissions
 *
 *	Turn the new style permissions object in into a compatibility
 *	object and store it into the 'out' pointer.
 */
 
void ipc64_perm_to_ipc_perm (struct ipc64_perm *in, struct ipc_perm *out)
{
	out->key	= in->key;
Andi Kleen's avatar
Andi Kleen committed
475 476 477 478
	SET_UID(out->uid, in->uid);
	SET_GID(out->gid, in->gid);
	SET_UID(out->cuid, in->cuid);
	SET_GID(out->cgid, in->cgid);
Linus Torvalds's avatar
Linus Torvalds committed
479 480 481 482
	out->mode	= in->mode;
	out->seq	= in->seq;
}

Andrew Morton's avatar
Andrew Morton committed
483
/*
484 485 486 487 488
 * So far only shm_get_stat() calls ipc_get() via shm_get(), so ipc_get()
 * is called with shm_ids.sem locked.  Since grow_ary() is also called with
 * shm_ids.sem down(for Shared Memory), there is no need to add read 
 * barriers here to gurantee the writes in grow_ary() are seen in order 
 * here (for Alpha).
Andrew Morton's avatar
Andrew Morton committed
489
 *
490 491 492
 * However ipc_get() itself does not necessary require ipc_ids.sem down. So
 * if in the future ipc_get() is used by other places without ipc_ids.sem
 * down, then ipc_get() needs read memery barriers as ipc_lock() does.
Andrew Morton's avatar
Andrew Morton committed
493 494 495 496 497
 */
struct kern_ipc_perm* ipc_get(struct ipc_ids* ids, int id)
{
	struct kern_ipc_perm* out;
	int lid = id % SEQ_MULTIPLIER;
498
	if(lid >= ids->entries->size)
Andrew Morton's avatar
Andrew Morton committed
499
		return NULL;
500
	out = ids->entries->p[lid];
Andrew Morton's avatar
Andrew Morton committed
501 502 503 504 505 506 507
	return out;
}

struct kern_ipc_perm* ipc_lock(struct ipc_ids* ids, int id)
{
	struct kern_ipc_perm* out;
	int lid = id % SEQ_MULTIPLIER;
508
	struct ipc_id_ary* entries;
Andrew Morton's avatar
Andrew Morton committed
509 510

	rcu_read_lock();
511 512
	entries = rcu_dereference(ids->entries);
	if(lid >= entries->size) {
Andrew Morton's avatar
Andrew Morton committed
513 514 515
		rcu_read_unlock();
		return NULL;
	}
516
	out = entries->p[lid];
Andrew Morton's avatar
Andrew Morton committed
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
	if(out == NULL) {
		rcu_read_unlock();
		return NULL;
	}
	spin_lock(&out->lock);
	
	/* ipc_rmid() may have already freed the ID while ipc_lock
	 * was spinning: here verify that the structure is still valid
	 */
	if (out->deleted) {
		spin_unlock(&out->lock);
		rcu_read_unlock();
		return NULL;
	}
	return out;
}

534 535 536 537 538 539
void ipc_lock_by_ptr(struct kern_ipc_perm *perm)
{
	rcu_read_lock();
	spin_lock(&perm->lock);
}

Andrew Morton's avatar
Andrew Morton committed
540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557
void ipc_unlock(struct kern_ipc_perm* perm)
{
	spin_unlock(&perm->lock);
	rcu_read_unlock();
}

int ipc_buildid(struct ipc_ids* ids, int id, int seq)
{
	return SEQ_MULTIPLIER*seq + id;
}

int ipc_checkid(struct ipc_ids* ids, struct kern_ipc_perm* ipcp, int uid)
{
	if(uid/SEQ_MULTIPLIER != ipcp->seq)
		return 1;
	return 0;
}

558 559
#ifdef __ARCH_WANT_IPC_PARSE_VERSION

Linus Torvalds's avatar
Linus Torvalds committed
560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579

/**
 *	ipc_parse_version	-	IPC call version
 *	@cmd: pointer to command
 *
 *	Return IPC_64 for new style IPC and IPC_OLD for old style IPC. 
 *	The cmd value is turned from an encoding command and version into
 *	just the command code.
 */
 
int ipc_parse_version (int *cmd)
{
	if (*cmd & IPC_64) {
		*cmd ^= IPC_64;
		return IPC_64;
	} else {
		return IPC_OLD;
	}
}

580
#endif /* __ARCH_WANT_IPC_PARSE_VERSION */