file.c 15.3 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6
/*
 * file.c - operations for regular (text) files.
 */

#include <linux/module.h>
#include <linux/kobject.h>
7
#include <linux/namei.h>
8
#include <linux/poll.h>
9
#include <linux/list.h>
10
#include <linux/mutex.h>
L
Linus Torvalds 已提交
11 12 13 14
#include <asm/uaccess.h>

#include "sysfs.h"

15
#define to_sattr(a) container_of(a,struct subsys_attribute, attr)
L
Linus Torvalds 已提交
16

M
Martin Waitz 已提交
17
/*
L
Linus Torvalds 已提交
18 19 20 21 22 23 24
 * Subsystem file operations.
 * These operations allow subsystems to have files that can be 
 * read/written. 
 */
static ssize_t 
subsys_attr_show(struct kobject * kobj, struct attribute * attr, char * page)
{
25
	struct kset *kset = to_kset(kobj);
L
Linus Torvalds 已提交
26
	struct subsys_attribute * sattr = to_sattr(attr);
27
	ssize_t ret = -EIO;
L
Linus Torvalds 已提交
28 29

	if (sattr->show)
30
		ret = sattr->show(kset, page);
L
Linus Torvalds 已提交
31 32 33 34 35 36 37
	return ret;
}

static ssize_t 
subsys_attr_store(struct kobject * kobj, struct attribute * attr, 
		  const char * page, size_t count)
{
38
	struct kset *kset = to_kset(kobj);
L
Linus Torvalds 已提交
39
	struct subsys_attribute * sattr = to_sattr(attr);
40
	ssize_t ret = -EIO;
L
Linus Torvalds 已提交
41 42

	if (sattr->store)
43
		ret = sattr->store(kset, page, count);
L
Linus Torvalds 已提交
44 45 46 47 48 49 50 51
	return ret;
}

static struct sysfs_ops subsys_sysfs_ops = {
	.show	= subsys_attr_show,
	.store	= subsys_attr_store,
};

T
Tejun Heo 已提交
52 53 54 55 56
struct sysfs_buffer {
	size_t			count;
	loff_t			pos;
	char			* page;
	struct sysfs_ops	* ops;
57
	struct mutex		mutex;
T
Tejun Heo 已提交
58 59 60
	int			needs_read_fill;
	int			event;
};
L
Linus Torvalds 已提交
61 62 63 64 65 66 67 68 69

/**
 *	fill_read_buffer - allocate and fill buffer from object.
 *	@dentry:	dentry pointer.
 *	@buffer:	data buffer for file.
 *
 *	Allocate @buffer->page, if it hasn't been already, then call the
 *	kobject's show() method to fill the buffer with this attribute's 
 *	data. 
70 71
 *	This is called only once, on the file's first read unless an error
 *	is returned.
L
Linus Torvalds 已提交
72 73 74
 */
static int fill_read_buffer(struct dentry * dentry, struct sysfs_buffer * buffer)
{
75 76
	struct sysfs_dirent *attr_sd = dentry->d_fsdata;
	struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj;
L
Linus Torvalds 已提交
77 78 79 80 81 82 83 84 85
	struct sysfs_ops * ops = buffer->ops;
	int ret = 0;
	ssize_t count;

	if (!buffer->page)
		buffer->page = (char *) get_zeroed_page(GFP_KERNEL);
	if (!buffer->page)
		return -ENOMEM;

86 87 88 89 90 91 92 93 94
	/* need attr_sd for attr and ops, its parent for kobj */
	if (!sysfs_get_active_two(attr_sd))
		return -ENODEV;

	buffer->event = atomic_read(&attr_sd->s_event);
	count = ops->show(kobj, attr_sd->s_elem.attr.attr, buffer->page);

	sysfs_put_active_two(attr_sd);

L
Linus Torvalds 已提交
95
	BUG_ON(count > (ssize_t)PAGE_SIZE);
96 97
	if (count >= 0) {
		buffer->needs_read_fill = 0;
L
Linus Torvalds 已提交
98
		buffer->count = count;
99
	} else {
L
Linus Torvalds 已提交
100
		ret = count;
101
	}
L
Linus Torvalds 已提交
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
	return ret;
}

/**
 *	sysfs_read_file - read an attribute. 
 *	@file:	file pointer.
 *	@buf:	buffer to fill.
 *	@count:	number of bytes to read.
 *	@ppos:	starting offset in file.
 *
 *	Userspace wants to read an attribute file. The attribute descriptor
 *	is in the file's ->d_fsdata. The target object is in the directory's
 *	->d_fsdata.
 *
 *	We call fill_read_buffer() to allocate and fill the buffer from the
 *	object's show() method exactly once (if the read is happening from
 *	the beginning of the file). That should fill the entire buffer with
 *	all the data the object has to offer for that attribute.
 *	We then call flush_read_buffer() to copy the buffer to userspace
 *	in the increments specified.
 */

static ssize_t
sysfs_read_file(struct file *file, char __user *buf, size_t count, loff_t *ppos)
{
	struct sysfs_buffer * buffer = file->private_data;
	ssize_t retval = 0;

130
	mutex_lock(&buffer->mutex);
L
Linus Torvalds 已提交
131
	if (buffer->needs_read_fill) {
T
Tejun Heo 已提交
132
		retval = fill_read_buffer(file->f_path.dentry,buffer);
133
		if (retval)
L
Linus Torvalds 已提交
134 135
			goto out;
	}
136 137
	pr_debug("%s: count = %zd, ppos = %lld, buf = %s\n",
		 __FUNCTION__, count, *ppos, buffer->page);
138 139
	retval = simple_read_from_buffer(buf, count, ppos, buffer->page,
					 buffer->count);
L
Linus Torvalds 已提交
140
out:
141
	mutex_unlock(&buffer->mutex);
L
Linus Torvalds 已提交
142 143 144 145 146 147
	return retval;
}

/**
 *	fill_write_buffer - copy buffer from userspace.
 *	@buffer:	data buffer for file.
148
 *	@buf:		data from user.
L
Linus Torvalds 已提交
149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165
 *	@count:		number of bytes in @userbuf.
 *
 *	Allocate @buffer->page if it hasn't been already, then
 *	copy the user-supplied buffer into it.
 */

static int 
fill_write_buffer(struct sysfs_buffer * buffer, const char __user * buf, size_t count)
{
	int error;

	if (!buffer->page)
		buffer->page = (char *)get_zeroed_page(GFP_KERNEL);
	if (!buffer->page)
		return -ENOMEM;

	if (count >= PAGE_SIZE)
166
		count = PAGE_SIZE - 1;
L
Linus Torvalds 已提交
167 168
	error = copy_from_user(buffer->page,buf,count);
	buffer->needs_read_fill = 1;
169 170 171
	/* if buf is assumed to contain a string, terminate it by \0,
	   so e.g. sscanf() can scan the string easily */
	buffer->page[count] = 0;
L
Linus Torvalds 已提交
172 173 174 175 176 177
	return error ? -EFAULT : count;
}


/**
 *	flush_write_buffer - push buffer to kobject.
M
Martin Waitz 已提交
178
 *	@dentry:	dentry to the attribute
L
Linus Torvalds 已提交
179
 *	@buffer:	data buffer for file.
M
Martin Waitz 已提交
180
 *	@count:		number of bytes
L
Linus Torvalds 已提交
181 182 183 184 185 186
 *
 *	Get the correct pointers for the kobject and the attribute we're
 *	dealing with, then call the store() method for the attribute, 
 *	passing the buffer that we acquired in fill_write_buffer().
 */

187
static int
L
Linus Torvalds 已提交
188 189
flush_write_buffer(struct dentry * dentry, struct sysfs_buffer * buffer, size_t count)
{
190
	struct sysfs_dirent *attr_sd = dentry->d_fsdata;
191
	struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj;
L
Linus Torvalds 已提交
192
	struct sysfs_ops * ops = buffer->ops;
193 194 195 196 197 198 199 200 201
	int rc;

	/* need attr_sd for attr and ops, its parent for kobj */
	if (!sysfs_get_active_two(attr_sd))
		return -ENODEV;

	rc = ops->store(kobj, attr_sd->s_elem.attr.attr, buffer->page, count);

	sysfs_put_active_two(attr_sd);
L
Linus Torvalds 已提交
202

203
	return rc;
L
Linus Torvalds 已提交
204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229
}


/**
 *	sysfs_write_file - write an attribute.
 *	@file:	file pointer
 *	@buf:	data to write
 *	@count:	number of bytes
 *	@ppos:	starting offset
 *
 *	Similar to sysfs_read_file(), though working in the opposite direction.
 *	We allocate and fill the data from the user in fill_write_buffer(),
 *	then push it to the kobject in flush_write_buffer().
 *	There is no easy way for us to know if userspace is only doing a partial
 *	write, so we don't support them. We expect the entire buffer to come
 *	on the first write. 
 *	Hint: if you're writing a value, first read the file, modify only the
 *	the value you're changing, then write entire buffer back. 
 */

static ssize_t
sysfs_write_file(struct file *file, const char __user *buf, size_t count, loff_t *ppos)
{
	struct sysfs_buffer * buffer = file->private_data;
	ssize_t len;

230
	mutex_lock(&buffer->mutex);
L
Linus Torvalds 已提交
231 232
	len = fill_write_buffer(buffer, buf, count);
	if (len > 0)
233
		len = flush_write_buffer(file->f_path.dentry, buffer, len);
L
Linus Torvalds 已提交
234 235
	if (len > 0)
		*ppos += len;
236
	mutex_unlock(&buffer->mutex);
L
Linus Torvalds 已提交
237 238 239
	return len;
}

240
static int sysfs_open_file(struct inode *inode, struct file *file)
L
Linus Torvalds 已提交
241
{
242
	struct sysfs_dirent *attr_sd = file->f_path.dentry->d_fsdata;
243
	struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj;
L
Linus Torvalds 已提交
244 245
	struct sysfs_buffer * buffer;
	struct sysfs_ops * ops = NULL;
246
	int error;
L
Linus Torvalds 已提交
247

248 249 250
	/* need attr_sd for attr and ops, its parent for kobj */
	if (!sysfs_get_active_two(attr_sd))
		return -ENODEV;
L
Linus Torvalds 已提交
251 252 253 254 255 256 257 258 259 260 261

	/* if the kobject has no ktype, then we assume that it is a subsystem
	 * itself, and use ops for it.
	 */
	if (kobj->kset && kobj->kset->ktype)
		ops = kobj->kset->ktype->sysfs_ops;
	else if (kobj->ktype)
		ops = kobj->ktype->sysfs_ops;
	else
		ops = &subsys_sysfs_ops;

T
Tejun Heo 已提交
262 263
	error = -EACCES;

L
Linus Torvalds 已提交
264 265 266 267
	/* No sysfs operations, either from having no subsystem,
	 * or the subsystem have no operations.
	 */
	if (!ops)
268
		goto err_out;
L
Linus Torvalds 已提交
269 270 271 272 273 274 275

	/* File needs write support.
	 * The inode's perms must say it's ok, 
	 * and we must have a store method.
	 */
	if (file->f_mode & FMODE_WRITE) {
		if (!(inode->i_mode & S_IWUGO) || !ops->store)
276
			goto err_out;
L
Linus Torvalds 已提交
277 278 279 280 281 282 283 284
	}

	/* File needs read support.
	 * The inode's perms must say it's ok, and we there
	 * must be a show method for it.
	 */
	if (file->f_mode & FMODE_READ) {
		if (!(inode->i_mode & S_IRUGO) || !ops->show)
285
			goto err_out;
L
Linus Torvalds 已提交
286 287 288 289 290
	}

	/* No error? Great, allocate a buffer for the file, and store it
	 * it in file->private_data for easy access.
	 */
291
	error = -ENOMEM;
292
	buffer = kzalloc(sizeof(struct sysfs_buffer), GFP_KERNEL);
293
	if (!buffer)
294
		goto err_out;
L
Linus Torvalds 已提交
295

296
	mutex_init(&buffer->mutex);
297 298 299 300 301 302 303 304 305
	buffer->needs_read_fill = 1;
	buffer->ops = ops;
	file->private_data = buffer;

	/* open succeeded, put active references and pin attr_sd */
	sysfs_put_active_two(attr_sd);
	sysfs_get(attr_sd);
	return 0;

306
 err_out:
307
	sysfs_put_active_two(attr_sd);
L
Linus Torvalds 已提交
308 309 310 311 312
	return error;
}

static int sysfs_release(struct inode * inode, struct file * filp)
{
313
	struct sysfs_dirent *attr_sd = filp->f_path.dentry->d_fsdata;
T
Tejun Heo 已提交
314
	struct sysfs_buffer *buffer = filp->private_data;
L
Linus Torvalds 已提交
315

316
	sysfs_put(attr_sd);
L
Linus Torvalds 已提交
317 318 319 320 321 322 323 324 325

	if (buffer) {
		if (buffer->page)
			free_page((unsigned long)buffer->page);
		kfree(buffer);
	}
	return 0;
}

326 327 328 329 330 331 332 333 334 335 336
/* Sysfs attribute files are pollable.  The idea is that you read
 * the content and then you use 'poll' or 'select' to wait for
 * the content to change.  When the content changes (assuming the
 * manager for the kobject supports notification), poll will
 * return POLLERR|POLLPRI, and select will return the fd whether
 * it is waiting for read, write, or exceptions.
 * Once poll/select indicates that the value has changed, you
 * need to close and re-open the file, as simply seeking and reading
 * again will not get new data, or reset the state of 'poll'.
 * Reminder: this only works for attributes which actively support
 * it, and it is not possible to test an attribute from userspace
337
 * to see if it supports poll (Neither 'poll' nor 'select' return
338 339 340 341 342
 * an appropriate error code).  When in doubt, set a suitable timeout value.
 */
static unsigned int sysfs_poll(struct file *filp, poll_table *wait)
{
	struct sysfs_buffer * buffer = filp->private_data;
343 344 345 346 347 348
	struct sysfs_dirent *attr_sd = filp->f_path.dentry->d_fsdata;
	struct kobject *kobj = attr_sd->s_parent->s_elem.dir.kobj;

	/* need parent for the kobj, grab both */
	if (!sysfs_get_active_two(attr_sd))
		goto trigger;
349 350 351

	poll_wait(filp, &kobj->poll, wait);

352 353 354 355
	sysfs_put_active_two(attr_sd);

	if (buffer->event != atomic_read(&attr_sd->s_event))
		goto trigger;
356

357 358 359 360 361
	return 0;

 trigger:
	buffer->needs_read_fill = 1;
	return POLLERR|POLLPRI;
362 363
}

364
void sysfs_notify(struct kobject *k, char *dir, char *attr)
365
{
366
	struct sysfs_dirent *sd = k->sd;
367

368 369 370 371 372 373 374 375
	mutex_lock(&sysfs_mutex);

	if (sd && dir)
		sd = sysfs_find_dirent(sd, dir);
	if (sd && attr)
		sd = sysfs_find_dirent(sd, attr);
	if (sd) {
		atomic_inc(&sd->s_event);
376 377
		wake_up_interruptible(&k->poll);
	}
378 379

	mutex_unlock(&sysfs_mutex);
380 381 382
}
EXPORT_SYMBOL_GPL(sysfs_notify);

383
const struct file_operations sysfs_file_operations = {
L
Linus Torvalds 已提交
384 385 386 387 388
	.read		= sysfs_read_file,
	.write		= sysfs_write_file,
	.llseek		= generic_file_llseek,
	.open		= sysfs_open_file,
	.release	= sysfs_release,
389
	.poll		= sysfs_poll,
L
Linus Torvalds 已提交
390 391 392
};


393 394
int sysfs_add_file(struct sysfs_dirent *dir_sd, const struct attribute *attr,
		   int type)
L
Linus Torvalds 已提交
395 396
{
	umode_t mode = (attr->mode & S_IALLUGO) | S_IFREG;
397
	struct sysfs_addrm_cxt acxt;
398
	struct sysfs_dirent *sd;
399
	int rc;
L
Linus Torvalds 已提交
400

401 402 403 404
	sd = sysfs_new_dirent(attr->name, mode, type);
	if (!sd)
		return -ENOMEM;
	sd->s_elem.attr.attr = (void *)attr;
L
Linus Torvalds 已提交
405

406
	sysfs_addrm_start(&acxt, dir_sd);
407 408
	rc = sysfs_add_one(&acxt, sd);
	sysfs_addrm_finish(&acxt);
409

410
	if (rc)
411
		sysfs_put(sd);
412

413
	return rc;
L
Linus Torvalds 已提交
414 415 416 417 418 419 420 421 422 423 424
}


/**
 *	sysfs_create_file - create an attribute file for an object.
 *	@kobj:	object we're creating for. 
 *	@attr:	atrribute descriptor.
 */

int sysfs_create_file(struct kobject * kobj, const struct attribute * attr)
{
425
	BUG_ON(!kobj || !kobj->sd || !attr);
L
Linus Torvalds 已提交
426

427
	return sysfs_add_file(kobj->sd, attr, SYSFS_KOBJ_ATTR);
L
Linus Torvalds 已提交
428 429 430 431

}


432 433 434 435 436 437 438 439 440
/**
 * sysfs_add_file_to_group - add an attribute file to a pre-existing group.
 * @kobj: object we're acting for.
 * @attr: attribute descriptor.
 * @group: group name.
 */
int sysfs_add_file_to_group(struct kobject *kobj,
		const struct attribute *attr, const char *group)
{
441
	struct sysfs_dirent *dir_sd;
442 443
	int error;

444 445 446 447 448 449 450
	dir_sd = sysfs_get_dirent(kobj->sd, group);
	if (!dir_sd)
		return -ENOENT;

	error = sysfs_add_file(dir_sd, attr, SYSFS_KOBJ_ATTR);
	sysfs_put(dir_sd);

451 452 453 454
	return error;
}
EXPORT_SYMBOL_GPL(sysfs_add_file_to_group);

455 456 457 458 459 460 461 462 463
/**
 * sysfs_chmod_file - update the modified mode value on an object attribute.
 * @kobj: object we're acting for.
 * @attr: attribute descriptor.
 * @mode: file permissions.
 *
 */
int sysfs_chmod_file(struct kobject *kobj, struct attribute *attr, mode_t mode)
{
464 465
	struct sysfs_dirent *victim_sd = NULL;
	struct dentry *victim = NULL;
466 467
	struct inode * inode;
	struct iattr newattrs;
468 469 470 471 472 473 474
	int rc;

	rc = -ENOENT;
	victim_sd = sysfs_get_dirent(kobj->sd, attr->name);
	if (!victim_sd)
		goto out;

475
	mutex_lock(&sysfs_rename_mutex);
476
	victim = sysfs_get_dentry(victim_sd);
477
	mutex_unlock(&sysfs_rename_mutex);
478 479 480 481
	if (IS_ERR(victim)) {
		rc = PTR_ERR(victim);
		victim = NULL;
		goto out;
482 483
	}

484
	inode = victim->d_inode;
485

486
	mutex_lock(&inode->i_mutex);
487

488 489 490
	newattrs.ia_mode = (mode & S_IALLUGO) | (inode->i_mode & ~S_IALLUGO);
	newattrs.ia_valid = ATTR_MODE | ATTR_CTIME;
	rc = notify_change(victim, &newattrs);
491 492 493 494 495 496 497

	if (rc == 0) {
		mutex_lock(&sysfs_mutex);
		victim_sd->s_mode = newattrs.ia_mode;
		mutex_unlock(&sysfs_mutex);
	}

498 499 500 501 502
	mutex_unlock(&inode->i_mutex);
 out:
	dput(victim);
	sysfs_put(victim_sd);
	return rc;
503 504 505 506
}
EXPORT_SYMBOL_GPL(sysfs_chmod_file);


L
Linus Torvalds 已提交
507 508 509 510 511 512 513 514 515 516
/**
 *	sysfs_remove_file - remove an object attribute.
 *	@kobj:	object we're acting for.
 *	@attr:	attribute descriptor.
 *
 *	Hash the attribute name and kill the victim.
 */

void sysfs_remove_file(struct kobject * kobj, const struct attribute * attr)
{
517
	sysfs_hash_and_remove(kobj->sd, attr->name);
L
Linus Torvalds 已提交
518 519 520
}


521 522 523 524 525 526 527 528 529
/**
 * sysfs_remove_file_from_group - remove an attribute file from a group.
 * @kobj: object we're acting for.
 * @attr: attribute descriptor.
 * @group: group name.
 */
void sysfs_remove_file_from_group(struct kobject *kobj,
		const struct attribute *attr, const char *group)
{
530
	struct sysfs_dirent *dir_sd;
531

532 533 534 535
	dir_sd = sysfs_get_dirent(kobj->sd, group);
	if (dir_sd) {
		sysfs_hash_and_remove(dir_sd, attr->name);
		sysfs_put(dir_sd);
536 537 538 539
	}
}
EXPORT_SYMBOL_GPL(sysfs_remove_file_from_group);

540 541 542 543
struct sysfs_schedule_callback_struct {
	struct kobject 		*kobj;
	void			(*func)(void *);
	void			*data;
544
	struct module		*owner;
545 546 547 548 549 550 551 552 553 554
	struct work_struct	work;
};

static void sysfs_schedule_callback_work(struct work_struct *work)
{
	struct sysfs_schedule_callback_struct *ss = container_of(work,
			struct sysfs_schedule_callback_struct, work);

	(ss->func)(ss->data);
	kobject_put(ss->kobj);
555
	module_put(ss->owner);
556 557 558 559 560 561 562 563
	kfree(ss);
}

/**
 * sysfs_schedule_callback - helper to schedule a callback for a kobject
 * @kobj: object we're acting for.
 * @func: callback function to invoke later.
 * @data: argument to pass to @func.
564
 * @owner: module owning the callback code
565 566 567 568 569 570 571 572 573 574 575 576
 *
 * sysfs attribute methods must not unregister themselves or their parent
 * kobject (which would amount to the same thing).  Attempts to do so will
 * deadlock, since unregistration is mutually exclusive with driver
 * callbacks.
 *
 * Instead methods can call this routine, which will attempt to allocate
 * and schedule a workqueue request to call back @func with @data as its
 * argument in the workqueue's process context.  @kobj will be pinned
 * until @func returns.
 *
 * Returns 0 if the request was submitted, -ENOMEM if storage could not
577
 * be allocated, -ENODEV if a reference to @owner isn't available.
578 579
 */
int sysfs_schedule_callback(struct kobject *kobj, void (*func)(void *),
580
		void *data, struct module *owner)
581 582 583
{
	struct sysfs_schedule_callback_struct *ss;

584 585
	if (!try_module_get(owner))
		return -ENODEV;
586
	ss = kmalloc(sizeof(*ss), GFP_KERNEL);
587 588
	if (!ss) {
		module_put(owner);
589
		return -ENOMEM;
590
	}
591 592 593 594
	kobject_get(kobj);
	ss->kobj = kobj;
	ss->func = func;
	ss->data = data;
595
	ss->owner = owner;
596 597 598 599 600 601
	INIT_WORK(&ss->work, sysfs_schedule_callback_work);
	schedule_work(&ss->work);
	return 0;
}
EXPORT_SYMBOL_GPL(sysfs_schedule_callback);

602

L
Linus Torvalds 已提交
603 604
EXPORT_SYMBOL_GPL(sysfs_create_file);
EXPORT_SYMBOL_GPL(sysfs_remove_file);