file.c 30.3 KB
Newer Older
1
2
/*
  FUSE: Filesystem in Userspace
3
  Copyright (C) 2001-2006  Miklos Szeredi <miklos@szeredi.hu>
4
5
6
7
8
9
10
11
12
13

  This program can be distributed under the terms of the GNU GPL.
  See the file COPYING.
*/

#include "fuse_i.h"

#include <linux/pagemap.h>
#include <linux/slab.h>
#include <linux/kernel.h>
Alexey Dobriyan's avatar
Alexey Dobriyan committed
14
#include <linux/sched.h>
15

16
static const struct file_operations fuse_direct_io_file_operations;
17

18
19
static int fuse_send_open(struct inode *inode, struct file *file, int isdir,
			  struct fuse_open_out *outargp)
20
21
22
{
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_open_in inarg;
23
24
25
	struct fuse_req *req;
	int err;

26
27
28
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
29
30

	memset(&inarg, 0, sizeof(inarg));
31
32
33
	inarg.flags = file->f_flags & ~(O_CREAT | O_EXCL | O_NOCTTY);
	if (!fc->atomic_o_trunc)
		inarg.flags &= ~O_TRUNC;
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
	req->in.h.opcode = isdir ? FUSE_OPENDIR : FUSE_OPEN;
	req->in.h.nodeid = get_node_id(inode);
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(*outargp);
	req->out.args[0].value = outargp;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);

	return err;
}

struct fuse_file *fuse_file_alloc(void)
{
	struct fuse_file *ff;
	ff = kmalloc(sizeof(struct fuse_file), GFP_KERNEL);
	if (ff) {
54
55
		ff->reserved_req = fuse_request_alloc();
		if (!ff->reserved_req) {
56
57
			kfree(ff);
			ff = NULL;
58
59
60
		} else {
			INIT_LIST_HEAD(&ff->write_entry);
			atomic_set(&ff->count, 0);
61
62
63
64
65
66
67
		}
	}
	return ff;
}

void fuse_file_free(struct fuse_file *ff)
{
68
	fuse_request_free(ff->reserved_req);
69
70
71
	kfree(ff);
}

72
73
74
75
76
77
static struct fuse_file *fuse_file_get(struct fuse_file *ff)
{
	atomic_inc(&ff->count);
	return ff;
}

Miklos Szeredi's avatar
Miklos Szeredi committed
78
79
static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req)
{
80
81
	dput(req->misc.release.dentry);
	mntput(req->misc.release.vfsmount);
Miklos Szeredi's avatar
Miklos Szeredi committed
82
83
84
	fuse_put_request(fc, req);
}

85
86
87
88
static void fuse_file_put(struct fuse_file *ff)
{
	if (atomic_dec_and_test(&ff->count)) {
		struct fuse_req *req = ff->reserved_req;
89
90
		struct inode *inode = req->misc.release.dentry->d_inode;
		struct fuse_conn *fc = get_fuse_conn(inode);
Miklos Szeredi's avatar
Miklos Szeredi committed
91
		req->end = fuse_release_end;
92
93
94
95
96
		request_send_background(fc, req);
		kfree(ff);
	}
}

97
98
99
100
101
102
void fuse_finish_open(struct inode *inode, struct file *file,
		      struct fuse_file *ff, struct fuse_open_out *outarg)
{
	if (outarg->open_flags & FOPEN_DIRECT_IO)
		file->f_op = &fuse_direct_io_file_operations;
	if (!(outarg->open_flags & FOPEN_KEEP_CACHE))
Miklos Szeredi's avatar
Miklos Szeredi committed
103
		invalidate_inode_pages2(inode->i_mapping);
104
	ff->fh = outarg->fh;
105
	file->private_data = fuse_file_get(ff);
106
107
108
109
}

int fuse_open_common(struct inode *inode, struct file *file, int isdir)
{
110
111
112
113
	struct fuse_open_out outarg;
	struct fuse_file *ff;
	int err;

Miklos Szeredi's avatar
Miklos Szeredi committed
114
115
116
117
	/* VFS checks this, but only _after_ ->open() */
	if (file->f_flags & O_DIRECT)
		return -EINVAL;

118
119
120
121
	err = generic_file_open(inode, file);
	if (err)
		return err;

122
	ff = fuse_file_alloc();
123
	if (!ff)
124
		return -ENOMEM;
125

126
127
128
129
130
131
132
	err = fuse_send_open(inode, file, isdir, &outarg);
	if (err)
		fuse_file_free(ff);
	else {
		if (isdir)
			outarg.open_flags &= ~FOPEN_DIRECT_IO;
		fuse_finish_open(inode, file, ff, &outarg);
133
134
135
136
137
	}

	return err;
}

138
void fuse_release_fill(struct fuse_file *ff, u64 nodeid, int flags, int opcode)
139
{
140
	struct fuse_req *req = ff->reserved_req;
141
	struct fuse_release_in *inarg = &req->misc.release.in;
142
143

	inarg->fh = ff->fh;
144
	inarg->flags = flags;
145
	req->in.h.opcode = opcode;
146
	req->in.h.nodeid = nodeid;
147
148
149
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(struct fuse_release_in);
	req->in.args[0].value = inarg;
150
151
152
153
154
155
}

int fuse_release_common(struct inode *inode, struct file *file, int isdir)
{
	struct fuse_file *ff = file->private_data;
	if (ff) {
156
		struct fuse_conn *fc = get_fuse_conn(inode);
157
		struct fuse_req *req = ff->reserved_req;
158

159
160
		fuse_release_fill(ff, get_node_id(inode), file->f_flags,
				  isdir ? FUSE_RELEASEDIR : FUSE_RELEASE);
161
162

		/* Hold vfsmount and dentry until release is finished */
163
164
		req->misc.release.vfsmount = mntget(file->f_path.mnt);
		req->misc.release.dentry = dget(file->f_path.dentry);
165
166
167
168

		spin_lock(&fc->lock);
		list_del(&ff->write_entry);
		spin_unlock(&fc->lock);
169
170
171
172
173
174
		/*
		 * Normally this will send the RELEASE request,
		 * however if some asynchronous READ or WRITE requests
		 * are outstanding, the sending will be delayed
		 */
		fuse_file_put(ff);
175
	}
176
177
178
179
180

	/* Return value is ignored by VFS */
	return 0;
}

181
182
183
184
185
186
187
188
189
190
static int fuse_open(struct inode *inode, struct file *file)
{
	return fuse_open_common(inode, file, 0);
}

static int fuse_release(struct inode *inode, struct file *file)
{
	return fuse_release_common(inode, file, 0);
}

191
/*
192
193
 * Scramble the ID space with XTEA, so that the value of the files_struct
 * pointer is not exposed to userspace.
194
 */
195
u64 fuse_lock_owner_id(struct fuse_conn *fc, fl_owner_t id)
196
{
197
198
199
200
201
202
203
204
205
206
207
208
209
210
	u32 *k = fc->scramble_key;
	u64 v = (unsigned long) id;
	u32 v0 = v;
	u32 v1 = v >> 32;
	u32 sum = 0;
	int i;

	for (i = 0; i < 32; i++) {
		v0 += ((v1 << 4 ^ v1 >> 5) + v1) ^ (sum + k[sum & 3]);
		sum += 0x9E3779B9;
		v1 += ((v0 << 4 ^ v0 >> 5) + v0) ^ (sum + k[sum>>11 & 3]);
	}

	return (u64) v0 + ((u64) v1 << 32);
211
212
}

Miklos Szeredi's avatar
Miklos Szeredi committed
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
/*
 * Check if page is under writeback
 *
 * This is currently done by walking the list of writepage requests
 * for the inode, which can be pretty inefficient.
 */
static bool fuse_page_is_writeback(struct inode *inode, pgoff_t index)
{
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_inode *fi = get_fuse_inode(inode);
	struct fuse_req *req;
	bool found = false;

	spin_lock(&fc->lock);
	list_for_each_entry(req, &fi->writepages, writepages_entry) {
		pgoff_t curr_index;

		BUG_ON(req->inode != inode);
		curr_index = req->misc.write.in.offset >> PAGE_CACHE_SHIFT;
		if (curr_index == index) {
			found = true;
			break;
		}
	}
	spin_unlock(&fc->lock);

	return found;
}

/*
 * Wait for page writeback to be completed.
 *
 * Since fuse doesn't rely on the VM writeback tracking, this has to
 * use some other means.
 */
static int fuse_wait_on_page_writeback(struct inode *inode, pgoff_t index)
{
	struct fuse_inode *fi = get_fuse_inode(inode);

	wait_event(fi->page_waitq, !fuse_page_is_writeback(inode, index));
	return 0;
}

256
static int fuse_flush(struct file *file, fl_owner_t id)
257
{
258
	struct inode *inode = file->f_path.dentry->d_inode;
259
260
261
262
263
264
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_file *ff = file->private_data;
	struct fuse_req *req;
	struct fuse_flush_in inarg;
	int err;

265
266
267
	if (is_bad_inode(inode))
		return -EIO;

268
269
270
	if (fc->no_flush)
		return 0;

271
	req = fuse_get_req_nofail(fc, file);
272
273
	memset(&inarg, 0, sizeof(inarg));
	inarg.fh = ff->fh;
274
	inarg.lock_owner = fuse_lock_owner_id(fc, id);
275
276
277
278
279
	req->in.h.opcode = FUSE_FLUSH;
	req->in.h.nodeid = get_node_id(inode);
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
280
	req->force = 1;
281
	request_send(fc, req);
282
283
284
285
286
287
288
289
290
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS) {
		fc->no_flush = 1;
		err = 0;
	}
	return err;
}

Miklos Szeredi's avatar
Miklos Szeredi committed
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
/*
 * Wait for all pending writepages on the inode to finish.
 *
 * This is currently done by blocking further writes with FUSE_NOWRITE
 * and waiting for all sent writes to complete.
 *
 * This must be called under i_mutex, otherwise the FUSE_NOWRITE usage
 * could conflict with truncation.
 */
static void fuse_sync_writes(struct inode *inode)
{
	fuse_set_nowrite(inode);
	fuse_release_nowrite(inode);
}

306
307
int fuse_fsync_common(struct file *file, struct dentry *de, int datasync,
		      int isdir)
308
309
310
311
312
313
314
315
{
	struct inode *inode = de->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_file *ff = file->private_data;
	struct fuse_req *req;
	struct fuse_fsync_in inarg;
	int err;

316
317
318
	if (is_bad_inode(inode))
		return -EIO;

319
	if ((!isdir && fc->no_fsync) || (isdir && fc->no_fsyncdir))
320
321
		return 0;

Miklos Szeredi's avatar
Miklos Szeredi committed
322
323
324
325
326
327
328
329
330
331
332
	/*
	 * Start writeback against all dirty pages of the inode, then
	 * wait for all outstanding writes, before sending the FSYNC
	 * request.
	 */
	err = write_inode_now(inode, 0);
	if (err)
		return err;

	fuse_sync_writes(inode);

333
334
335
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
336
337
338
339

	memset(&inarg, 0, sizeof(inarg));
	inarg.fh = ff->fh;
	inarg.fsync_flags = datasync ? 1 : 0;
340
	req->in.h.opcode = isdir ? FUSE_FSYNCDIR : FUSE_FSYNC;
341
342
343
344
345
346
347
348
	req->in.h.nodeid = get_node_id(inode);
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS) {
349
350
351
352
		if (isdir)
			fc->no_fsyncdir = 1;
		else
			fc->no_fsync = 1;
353
354
355
356
357
		err = 0;
	}
	return err;
}

358
359
360
361
362
static int fuse_fsync(struct file *file, struct dentry *de, int datasync)
{
	return fuse_fsync_common(file, de, datasync, 0);
}

363
void fuse_read_fill(struct fuse_req *req, struct file *file,
364
		    struct inode *inode, loff_t pos, size_t count, int opcode)
365
{
366
	struct fuse_read_in *inarg = &req->misc.read_in;
367
	struct fuse_file *ff = file->private_data;
368

369
370
371
	inarg->fh = ff->fh;
	inarg->offset = pos;
	inarg->size = count;
372
	inarg->flags = file->f_flags;
373
	req->in.h.opcode = opcode;
374
375
376
	req->in.h.nodeid = get_node_id(inode);
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(struct fuse_read_in);
377
	req->in.args[0].value = inarg;
378
379
380
381
382
383
	req->out.argpages = 1;
	req->out.argvar = 1;
	req->out.numargs = 1;
	req->out.args[0].size = count;
}

384
static size_t fuse_send_read(struct fuse_req *req, struct file *file,
385
386
			     struct inode *inode, loff_t pos, size_t count,
			     fl_owner_t owner)
387
{
388
	struct fuse_conn *fc = get_fuse_conn(inode);
389

390
	fuse_read_fill(req, file, inode, pos, count, FUSE_READ);
391
392
393
394
395
396
	if (owner != NULL) {
		struct fuse_read_in *inarg = &req->misc.read_in;

		inarg->read_flags |= FUSE_READ_LOCKOWNER;
		inarg->lock_owner = fuse_lock_owner_id(fc, owner);
	}
397
398
	request_send(fc, req);
	return req->out.args[0].size;
399
400
}

401
402
403
404
static int fuse_readpage(struct file *file, struct page *page)
{
	struct inode *inode = page->mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
405
406
407
408
409
410
411
	struct fuse_req *req;
	int err;

	err = -EIO;
	if (is_bad_inode(inode))
		goto out;

Miklos Szeredi's avatar
Miklos Szeredi committed
412
413
414
415
416
417
418
	/*
	 * Page writeback can extend beyond the liftime of the
	 * page-cache page, so make sure we read a properly synced
	 * page.
	 */
	fuse_wait_on_page_writeback(inode, page->index);

419
420
421
	req = fuse_get_req(fc);
	err = PTR_ERR(req);
	if (IS_ERR(req))
422
423
424
425
426
		goto out;

	req->out.page_zeroing = 1;
	req->num_pages = 1;
	req->pages[0] = page;
427
428
	fuse_send_read(req, file, inode, page_offset(page), PAGE_CACHE_SIZE,
		       NULL);
429
430
431
432
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err)
		SetPageUptodate(page);
433
	fuse_invalidate_attr(inode); /* atime changed */
434
435
436
437
438
 out:
	unlock_page(page);
	return err;
}

439
static void fuse_readpages_end(struct fuse_conn *fc, struct fuse_req *req)
440
{
441
442
443
444
	int i;

	fuse_invalidate_attr(req->pages[0]->mapping->host); /* atime changed */

445
446
447
448
	for (i = 0; i < req->num_pages; i++) {
		struct page *page = req->pages[i];
		if (!req->out.h.error)
			SetPageUptodate(page);
449
450
		else
			SetPageError(page);
451
452
		unlock_page(page);
	}
453
454
	if (req->ff)
		fuse_file_put(req->ff);
455
456
457
	fuse_put_request(fc, req);
}

458
static void fuse_send_readpages(struct fuse_req *req, struct file *file,
459
460
461
462
463
464
				struct inode *inode)
{
	struct fuse_conn *fc = get_fuse_conn(inode);
	loff_t pos = page_offset(req->pages[0]);
	size_t count = req->num_pages << PAGE_CACHE_SHIFT;
	req->out.page_zeroing = 1;
465
	fuse_read_fill(req, file, inode, pos, count, FUSE_READ);
466
	if (fc->async_read) {
467
		struct fuse_file *ff = file->private_data;
468
		req->ff = fuse_file_get(ff);
469
470
471
472
473
474
		req->end = fuse_readpages_end;
		request_send_background(fc, req);
	} else {
		request_send(fc, req);
		fuse_readpages_end(fc, req);
	}
475
476
}

477
struct fuse_fill_data {
478
	struct fuse_req *req;
479
	struct file *file;
480
481
482
483
484
	struct inode *inode;
};

static int fuse_readpages_fill(void *_data, struct page *page)
{
485
	struct fuse_fill_data *data = _data;
486
487
488
489
	struct fuse_req *req = data->req;
	struct inode *inode = data->inode;
	struct fuse_conn *fc = get_fuse_conn(inode);

Miklos Szeredi's avatar
Miklos Szeredi committed
490
491
	fuse_wait_on_page_writeback(inode, page->index);

492
493
494
495
	if (req->num_pages &&
	    (req->num_pages == FUSE_MAX_PAGES_PER_REQ ||
	     (req->num_pages + 1) * PAGE_CACHE_SIZE > fc->max_read ||
	     req->pages[req->num_pages - 1]->index + 1 != page->index)) {
496
		fuse_send_readpages(req, data->file, inode);
497
498
		data->req = req = fuse_get_req(fc);
		if (IS_ERR(req)) {
499
			unlock_page(page);
500
			return PTR_ERR(req);
501
502
503
504
505
506
507
508
509
510
511
512
		}
	}
	req->pages[req->num_pages] = page;
	req->num_pages ++;
	return 0;
}

static int fuse_readpages(struct file *file, struct address_space *mapping,
			  struct list_head *pages, unsigned nr_pages)
{
	struct inode *inode = mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
513
	struct fuse_fill_data data;
514
	int err;
515

516
	err = -EIO;
517
	if (is_bad_inode(inode))
518
		goto out;
519

520
	data.file = file;
521
	data.inode = inode;
522
	data.req = fuse_get_req(fc);
523
	err = PTR_ERR(data.req);
524
	if (IS_ERR(data.req))
525
		goto out;
526
527

	err = read_cache_pages(mapping, pages, fuse_readpages_fill, &data);
528
529
	if (!err) {
		if (data.req->num_pages)
530
			fuse_send_readpages(data.req, file, inode);
531
532
533
		else
			fuse_put_request(fc, data.req);
	}
534
out:
535
	return err;
536
537
}

Miklos Szeredi's avatar
Miklos Szeredi committed
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
static ssize_t fuse_file_aio_read(struct kiocb *iocb, const struct iovec *iov,
				  unsigned long nr_segs, loff_t pos)
{
	struct inode *inode = iocb->ki_filp->f_mapping->host;

	if (pos + iov_length(iov, nr_segs) > i_size_read(inode)) {
		int err;
		/*
		 * If trying to read past EOF, make sure the i_size
		 * attribute is up-to-date.
		 */
		err = fuse_update_attributes(inode, NULL, iocb->ki_filp, NULL);
		if (err)
			return err;
	}

	return generic_file_aio_read(iocb, iov, nr_segs, pos);
}

557
static void fuse_write_fill(struct fuse_req *req, struct file *file,
Miklos Szeredi's avatar
Miklos Szeredi committed
558
559
			    struct fuse_file *ff, struct inode *inode,
			    loff_t pos, size_t count, int writepage)
560
{
561
	struct fuse_conn *fc = get_fuse_conn(inode);
562
563
	struct fuse_write_in *inarg = &req->misc.write.in;
	struct fuse_write_out *outarg = &req->misc.write.out;
564

565
566
567
568
569
	memset(inarg, 0, sizeof(struct fuse_write_in));
	inarg->fh = ff->fh;
	inarg->offset = pos;
	inarg->size = count;
	inarg->write_flags = writepage ? FUSE_WRITE_CACHE : 0;
Miklos Szeredi's avatar
Miklos Szeredi committed
570
	inarg->flags = file ? file->f_flags : 0;
571
572
573
574
	req->in.h.opcode = FUSE_WRITE;
	req->in.h.nodeid = get_node_id(inode);
	req->in.argpages = 1;
	req->in.numargs = 2;
575
576
577
578
	if (fc->minor < 9)
		req->in.args[0].size = FUSE_COMPAT_WRITE_IN_SIZE;
	else
		req->in.args[0].size = sizeof(struct fuse_write_in);
579
	req->in.args[0].value = inarg;
580
581
582
	req->in.args[1].size = count;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(struct fuse_write_out);
583
584
585
586
	req->out.args[0].value = outarg;
}

static size_t fuse_send_write(struct fuse_req *req, struct file *file,
587
588
			      struct inode *inode, loff_t pos, size_t count,
			      fl_owner_t owner)
589
590
{
	struct fuse_conn *fc = get_fuse_conn(inode);
Miklos Szeredi's avatar
Miklos Szeredi committed
591
	fuse_write_fill(req, file, file->private_data, inode, pos, count, 0);
592
593
594
595
596
	if (owner != NULL) {
		struct fuse_write_in *inarg = &req->misc.write.in;
		inarg->write_flags |= FUSE_WRITE_LOCKOWNER;
		inarg->lock_owner = fuse_lock_owner_id(fc, owner);
	}
597
	request_send(fc, req);
598
	return req->misc.write.out.size;
599
600
}

Nick Piggin's avatar
Nick Piggin committed
601
602
603
static int fuse_write_begin(struct file *file, struct address_space *mapping,
			loff_t pos, unsigned len, unsigned flags,
			struct page **pagep, void **fsdata)
604
{
Nick Piggin's avatar
Nick Piggin committed
605
606
607
608
609
	pgoff_t index = pos >> PAGE_CACHE_SHIFT;

	*pagep = __grab_cache_page(mapping, index);
	if (!*pagep)
		return -ENOMEM;
610
611
612
	return 0;
}

613
614
615
616
617
618
619
620
621
622
623
624
static void fuse_write_update_size(struct inode *inode, loff_t pos)
{
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_inode *fi = get_fuse_inode(inode);

	spin_lock(&fc->lock);
	fi->attr_version = ++fc->attr_version;
	if (pos > inode->i_size)
		i_size_write(inode, pos);
	spin_unlock(&fc->lock);
}

Nick Piggin's avatar
Nick Piggin committed
625
626
static int fuse_buffered_write(struct file *file, struct inode *inode,
			       loff_t pos, unsigned count, struct page *page)
627
628
{
	int err;
629
	size_t nres;
630
	struct fuse_conn *fc = get_fuse_conn(inode);
Nick Piggin's avatar
Nick Piggin committed
631
	unsigned offset = pos & (PAGE_CACHE_SIZE - 1);
632
633
634
635
636
	struct fuse_req *req;

	if (is_bad_inode(inode))
		return -EIO;

Miklos Szeredi's avatar
Miklos Szeredi committed
637
638
639
640
641
642
	/*
	 * Make sure writepages on the same page are not mixed up with
	 * plain writes.
	 */
	fuse_wait_on_page_writeback(inode, page->index);

643
644
645
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
646
647
648
649

	req->num_pages = 1;
	req->pages[0] = page;
	req->page_offset = offset;
650
	nres = fuse_send_write(req, file, inode, pos, count, NULL);
651
652
	err = req->out.h.error;
	fuse_put_request(fc, req);
Nick Piggin's avatar
Nick Piggin committed
653
	if (!err && !nres)
654
655
		err = -EIO;
	if (!err) {
Nick Piggin's avatar
Nick Piggin committed
656
		pos += nres;
657
		fuse_write_update_size(inode, pos);
Nick Piggin's avatar
Nick Piggin committed
658
		if (count == PAGE_CACHE_SIZE)
659
			SetPageUptodate(page);
660
661
	}
	fuse_invalidate_attr(inode);
Nick Piggin's avatar
Nick Piggin committed
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
	return err ? err : nres;
}

static int fuse_write_end(struct file *file, struct address_space *mapping,
			loff_t pos, unsigned len, unsigned copied,
			struct page *page, void *fsdata)
{
	struct inode *inode = mapping->host;
	int res = 0;

	if (copied)
		res = fuse_buffered_write(file, inode, pos, copied, page);

	unlock_page(page);
	page_cache_release(page);
	return res;
678
679
}

Miklos Szeredi's avatar
Miklos Szeredi committed
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
static void fuse_release_user_pages(struct fuse_req *req, int write)
{
	unsigned i;

	for (i = 0; i < req->num_pages; i++) {
		struct page *page = req->pages[i];
		if (write)
			set_page_dirty_lock(page);
		put_page(page);
	}
}

static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf,
			       unsigned nbytes, int write)
{
	unsigned long user_addr = (unsigned long) buf;
	unsigned offset = user_addr & ~PAGE_MASK;
	int npages;

	/* This doesn't work with nfsd */
	if (!current->mm)
		return -EPERM;

	nbytes = min(nbytes, (unsigned) FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT);
	npages = (nbytes + offset + PAGE_SIZE - 1) >> PAGE_SHIFT;
705
	npages = min(max(npages, 1), FUSE_MAX_PAGES_PER_REQ);
Miklos Szeredi's avatar
Miklos Szeredi committed
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
	down_read(&current->mm->mmap_sem);
	npages = get_user_pages(current, current->mm, user_addr, npages, write,
				0, req->pages, NULL);
	up_read(&current->mm->mmap_sem);
	if (npages < 0)
		return npages;

	req->num_pages = npages;
	req->page_offset = offset;
	return 0;
}

static ssize_t fuse_direct_io(struct file *file, const char __user *buf,
			      size_t count, loff_t *ppos, int write)
{
721
	struct inode *inode = file->f_path.dentry->d_inode;
Miklos Szeredi's avatar
Miklos Szeredi committed
722
723
724
725
	struct fuse_conn *fc = get_fuse_conn(inode);
	size_t nmax = write ? fc->max_write : fc->max_read;
	loff_t pos = *ppos;
	ssize_t res = 0;
726
727
728
729
730
	struct fuse_req *req;

	if (is_bad_inode(inode))
		return -EIO;

731
732
733
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
Miklos Szeredi's avatar
Miklos Szeredi committed
734
735
736
737
738
739
740
741
742

	while (count) {
		size_t nres;
		size_t nbytes = min(count, nmax);
		int err = fuse_get_user_pages(req, buf, nbytes, !write);
		if (err) {
			res = err;
			break;
		}
743
744
		nbytes = (req->num_pages << PAGE_SHIFT) - req->page_offset;
		nbytes = min(count, nbytes);
Miklos Szeredi's avatar
Miklos Szeredi committed
745
		if (write)
746
747
			nres = fuse_send_write(req, file, inode, pos, nbytes,
					       current->files);
Miklos Szeredi's avatar
Miklos Szeredi committed
748
		else
749
750
			nres = fuse_send_read(req, file, inode, pos, nbytes,
					      current->files);
Miklos Szeredi's avatar
Miklos Szeredi committed
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
		fuse_release_user_pages(req, !write);
		if (req->out.h.error) {
			if (!res)
				res = req->out.h.error;
			break;
		} else if (nres > nbytes) {
			res = -EIO;
			break;
		}
		count -= nres;
		res += nres;
		pos += nres;
		buf += nres;
		if (nres != nbytes)
			break;
766
767
768
769
770
771
		if (count) {
			fuse_put_request(fc, req);
			req = fuse_get_req(fc);
			if (IS_ERR(req))
				break;
		}
Miklos Szeredi's avatar
Miklos Szeredi committed
772
773
774
	}
	fuse_put_request(fc, req);
	if (res > 0) {
775
776
		if (write)
			fuse_write_update_size(inode, pos);
Miklos Szeredi's avatar
Miklos Szeredi committed
777
		*ppos = pos;
778
779
	}
	fuse_invalidate_attr(inode);
Miklos Szeredi's avatar
Miklos Szeredi committed
780
781
782
783
784
785
786
787
788
789
790
791
792

	return res;
}

static ssize_t fuse_direct_read(struct file *file, char __user *buf,
				     size_t count, loff_t *ppos)
{
	return fuse_direct_io(file, buf, count, ppos, 0);
}

static ssize_t fuse_direct_write(struct file *file, const char __user *buf,
				 size_t count, loff_t *ppos)
{
793
	struct inode *inode = file->f_path.dentry->d_inode;
Miklos Szeredi's avatar
Miklos Szeredi committed
794
795
	ssize_t res;
	/* Don't allow parallel writes to the same file */
796
	mutex_lock(&inode->i_mutex);
797
798
799
	res = generic_write_checks(file, ppos, &count, 0);
	if (!res)
		res = fuse_direct_io(file, buf, count, ppos, 1);
800
	mutex_unlock(&inode->i_mutex);
Miklos Szeredi's avatar
Miklos Szeredi committed
801
802
803
	return res;
}

Miklos Szeredi's avatar
Miklos Szeredi committed
804
static void fuse_writepage_free(struct fuse_conn *fc, struct fuse_req *req)
805
{
Miklos Szeredi's avatar
Miklos Szeredi committed
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
	__free_page(req->pages[0]);
	fuse_file_put(req->ff);
	fuse_put_request(fc, req);
}

static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req)
{
	struct inode *inode = req->inode;
	struct fuse_inode *fi = get_fuse_inode(inode);
	struct backing_dev_info *bdi = inode->i_mapping->backing_dev_info;

	list_del(&req->writepages_entry);
	dec_bdi_stat(bdi, BDI_WRITEBACK);
	dec_zone_page_state(req->pages[0], NR_WRITEBACK_TEMP);
	bdi_writeout_inc(bdi);
	wake_up(&fi->page_waitq);
}

/* Called under fc->lock, may release and reacquire it */
static void fuse_send_writepage(struct fuse_conn *fc, struct fuse_req *req)
{
	struct fuse_inode *fi = get_fuse_inode(req->inode);
	loff_t size = i_size_read(req->inode);
	struct fuse_write_in *inarg = &req->misc.write.in;

	if (!fc->connected)
		goto out_free;

	if (inarg->offset + PAGE_CACHE_SIZE <= size) {
		inarg->size = PAGE_CACHE_SIZE;
	} else if (inarg->offset < size) {
		inarg->size = size & (PAGE_CACHE_SIZE - 1);
	} else {
		/* Got truncated off completely */
		goto out_free;
841
	}
Miklos Szeredi's avatar
Miklos Szeredi committed
842
843
844
845
846
847
848
849
850
851
852

	req->in.args[1].size = inarg->size;
	fi->writectr++;
	request_send_background_locked(fc, req);
	return;

 out_free:
	fuse_writepage_finish(fc, req);
	spin_unlock(&fc->lock);
	fuse_writepage_free(fc, req);
	spin_lock(&fc->lock);
853
854
}

Miklos Szeredi's avatar
Miklos Szeredi committed
855
856
857
858
859
860
861
/*
 * If fi->writectr is positive (no truncate or fsync going on) send
 * all queued writepage requests.
 *
 * Called with fc->lock
 */
void fuse_flush_writepages(struct inode *inode)
862
{
Miklos Szeredi's avatar
Miklos Szeredi committed
863
864
865
866
867
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
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
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
1014
1015
1016
1017
1018
1019
1020
1021
1022
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_inode *fi = get_fuse_inode(inode);
	struct fuse_req *req;

	while (fi->writectr >= 0 && !list_empty(&fi->queued_writes)) {
		req = list_entry(fi->queued_writes.next, struct fuse_req, list);
		list_del_init(&req->list);
		fuse_send_writepage(fc, req);
	}
}

static void fuse_writepage_end(struct fuse_conn *fc, struct fuse_req *req)
{
	struct inode *inode = req->inode;
	struct fuse_inode *fi = get_fuse_inode(inode);

	mapping_set_error(inode->i_mapping, req->out.h.error);
	spin_lock(&fc->lock);
	fi->writectr--;
	fuse_writepage_finish(fc, req);
	spin_unlock(&fc->lock);
	fuse_writepage_free(fc, req);
}

static int fuse_writepage_locked(struct page *page)
{
	struct address_space *mapping = page->mapping;
	struct inode *inode = mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_inode *fi = get_fuse_inode(inode);
	struct fuse_req *req;
	struct fuse_file *ff;
	struct page *tmp_page;

	set_page_writeback(page);

	req = fuse_request_alloc_nofs();
	if (!req)
		goto err;

	tmp_page = alloc_page(GFP_NOFS | __GFP_HIGHMEM);
	if (!tmp_page)
		goto err_free;

	spin_lock(&fc->lock);
	BUG_ON(list_empty(&fi->write_files));
	ff = list_entry(fi->write_files.next, struct fuse_file, write_entry);
	req->ff = fuse_file_get(ff);
	spin_unlock(&fc->lock);

	fuse_write_fill(req, NULL, ff, inode, page_offset(page), 0, 1);

	copy_highpage(tmp_page, page);
	req->num_pages = 1;
	req->pages[0] = tmp_page;
	req->page_offset = 0;
	req->end = fuse_writepage_end;
	req->inode = inode;

	inc_bdi_stat(mapping->backing_dev_info, BDI_WRITEBACK);
	inc_zone_page_state(tmp_page, NR_WRITEBACK_TEMP);
	end_page_writeback(page);

	spin_lock(&fc->lock);
	list_add(&req->writepages_entry, &fi->writepages);
	list_add_tail(&req->list, &fi->queued_writes);
	fuse_flush_writepages(inode);
	spin_unlock(&fc->lock);

	return 0;

err_free:
	fuse_request_free(req);
err:
	end_page_writeback(page);
	return -ENOMEM;
}

static int fuse_writepage(struct page *page, struct writeback_control *wbc)
{
	int err;

	err = fuse_writepage_locked(page);
	unlock_page(page);

	return err;
}

static int fuse_launder_page(struct page *page)
{
	int err = 0;
	if (clear_page_dirty_for_io(page)) {
		struct inode *inode = page->mapping->host;
		err = fuse_writepage_locked(page);
		if (!err)
			fuse_wait_on_page_writeback(inode, page->index);
	}
	return err;
}

/*
 * Write back dirty pages now, because there may not be any suitable
 * open files later
 */
static void fuse_vma_close(struct vm_area_struct *vma)
{
	filemap_write_and_wait(vma->vm_file->f_mapping);
}

/*
 * Wait for writeback against this page to complete before allowing it
 * to be marked dirty again, and hence written back again, possibly
 * before the previous writepage completed.
 *
 * Block here, instead of in ->writepage(), so that the userspace fs
 * can only block processes actually operating on the filesystem.
 *
 * Otherwise unprivileged userspace fs would be able to block
 * unrelated:
 *
 * - page migration
 * - sync(2)
 * - try_to_free_pages() with order > PAGE_ALLOC_COSTLY_ORDER
 */
static int fuse_page_mkwrite(struct vm_area_struct *vma, struct page *page)
{
	/*
	 * Don't use page->mapping as it may become NULL from a
	 * concurrent truncate.
	 */
	struct inode *inode = vma->vm_file->f_mapping->host;

	fuse_wait_on_page_writeback(inode, page->index);
	return 0;
}

static struct vm_operations_struct fuse_file_vm_ops = {
	.close		= fuse_vma_close,
	.fault		= filemap_fault,
	.page_mkwrite	= fuse_page_mkwrite,
};

static int fuse_file_mmap(struct file *file, struct vm_area_struct *vma)
{
	if ((vma->vm_flags & VM_SHARED) && (vma->vm_flags & VM_MAYWRITE)) {
		struct inode *inode = file->f_dentry->d_inode;
		struct fuse_conn *fc = get_fuse_conn(inode);
		struct fuse_inode *fi = get_fuse_inode(inode);
		struct fuse_file *ff = file->private_data;
		/*
		 * file may be written through mmap, so chain it onto the
		 * inodes's write_file list
		 */
		spin_lock(&fc->lock);
		if (list_empty(&ff->write_entry))
			list_add(&ff->write_entry, &fi->write_files);
		spin_unlock(&fc->lock);
	}
	file_accessed(file);
	vma->vm_ops = &fuse_file_vm_ops;
1023
1024
1025
	return 0;
}

1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
static int convert_fuse_file_lock(const struct fuse_file_lock *ffl,
				  struct file_lock *fl)
{
	switch (ffl->type) {
	case F_UNLCK:
		break;

	case F_RDLCK:
	case F_WRLCK:
		if (ffl->start > OFFSET_MAX || ffl->end > OFFSET_MAX ||
		    ffl->end < ffl->start)
			return -EIO;

		fl->fl_start = ffl->start;
		fl->fl_end = ffl->end;
		fl->fl_pid = ffl->pid;
		break;

	default:
		return -EIO;
	}
	fl->fl_type = ffl->type;
	return 0;
}

static void fuse_lk_fill(struct fuse_req *req, struct file *file,
1052
1053
			 const struct file_lock *fl, int opcode, pid_t pid,
			 int flock)
1054
{
1055
	struct inode *inode = file->f_path.dentry->d_inode;
1056
	struct fuse_conn *fc = get_fuse_conn(inode);
1057
1058
1059
1060
	struct fuse_file *ff = file->private_data;
	struct fuse_lk_in *arg = &req->misc.lk_in;

	arg->fh = ff->fh;
1061
	arg->owner = fuse_lock_owner_id(fc, fl->fl_owner);
1062
1063
1064
1065
	arg->lk.start = fl->fl_start;
	arg->lk.end = fl->fl_end;
	arg->lk.type = fl->fl_type;
	arg->lk.pid = pid;
1066
1067
	if (flock)
		arg->lk_flags |= FUSE_LK_FLOCK;
1068
1069
1070
1071
1072
1073
1074
1075
1076
	req->in.h.opcode = opcode;
	req->in.h.nodeid = get_node_id(inode);
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(*arg);
	req->in.args[0].value = arg;
}

static int fuse_getlk(struct file *file, struct file_lock *fl)
{
1077
	struct inode *inode = file->f_path.dentry->d_inode;
1078
1079
1080
1081
1082
1083
1084
1085
1086
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	struct fuse_lk_out outarg;
	int err;

	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);

1087
	fuse_lk_fill(req, file, fl, FUSE_GETLK, 0, 0);
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(outarg);
	req->out.args[0].value = &outarg;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err)
		err = convert_fuse_file_lock(&outarg.lk, fl);

	return err;
}

1100
static int fuse_setlk(struct file *file, struct file_lock *fl, int flock)
1101
{
1102
	struct inode *inode = file->f_path.dentry->d_inode;
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	int opcode = (fl->fl_flags & FL_SLEEP) ? FUSE_SETLKW : FUSE_SETLK;
	pid_t pid = fl->fl_type != F_UNLCK ? current->tgid : 0;
	int err;

	/* Unlock on close is handled by the flush method */
	if (fl->fl_flags & FL_CLOSE)
		return 0;

	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);

1117
	fuse_lk_fill(req, file, fl, opcode, pid, flock);
1118
1119
	request_send(fc, req);
	err = req->out.h.error;
1120
1121
1122
	/* locking is restartable */
	if (err == -EINTR)
		err = -ERESTARTSYS;
1123
1124
1125
1126
1127
1128
	fuse_put_request(fc, req);
	return err;
}

static int fuse_file_lock(struct file *file, int cmd, struct file_lock *fl)
{
1129
	struct inode *inode = file->f_path.dentry->d_inode;
1130
1131
1132
1133
1134
	struct fuse_conn *fc = get_fuse_conn(inode);
	int err;

	if (cmd == F_GETLK) {
		if (fc->no_lock) {
1135
			posix_test_lock(file, fl);
1136
1137
1138
1139
1140
1141
1142
			err = 0;
		} else
			err = fuse_getlk(file, fl);
	} else {
		if (fc->no_lock)
			err = posix_lock_file_wait(file, fl);
		else
1143
			err = fuse_setlk(file, fl, 0);
1144
1145
1146
1147
	}
	return err;
}

1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
static int fuse_file_flock(struct file *file, int cmd, struct file_lock *fl)
{
	struct inode *inode = file->f_path.dentry->d_inode;
	struct fuse_conn *fc = get_fuse_conn(inode);
	int err;

	if (fc->no_lock) {
		err = flock_lock_file_wait(file, fl);
	} else {
		/* emulate flock with POSIX locks */
		fl->fl_owner = (fl_owner_t) file;
		err = fuse_setlk(file, fl, 1);
	}

	return err;
}

1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
static sector_t fuse_bmap(struct address_space *mapping, sector_t block)
{
	struct inode *inode = mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_req *req;
	struct fuse_bmap_in inarg;
	struct fuse_bmap_out outarg;
	int err;

	if (!inode->i_sb->s_bdev || fc->no_bmap)
		return 0;

	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return 0;

	memset(&inarg, 0, sizeof(inarg));
	inarg.block = block;
	inarg.blocksize = inode->i_sb->s_blocksize;
	req->in.h.opcode = FUSE_BMAP;
	req->in.h.nodeid = get_node_id(inode);
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(outarg);
	req->out.args[0].value = &outarg;
	request_send(fc, req);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS)
		fc->no_bmap = 1;

	return err ? 0 : outarg.block;
}

1201
static const struct file_operations fuse_file_operations = {
1202
	.llseek		= generic_file_llseek,
1203
	.read		= do_sync_read,
Miklos Szeredi's avatar
Miklos Szeredi committed
1204
	.aio_read	= fuse_file_aio_read,
1205
1206
	.write		= do_sync_write,
	.aio_write	= generic_file_aio_write,
1207
1208
1209
1210
1211
	.mmap		= fuse_file_mmap,
	.open		= fuse_open,
	.flush		= fuse_flush,
	.release	= fuse_release,
	.fsync		= fuse_fsync,
1212
	.lock		= fuse_file_lock,
1213
	.flock		= fuse_file_flock,
1214
	.splice_read	= generic_file_splice_read,
1215
1216
};

1217
static const struct file_operations fuse_direct_io_file_operations = {
Miklos Szeredi's avatar
Miklos Szeredi committed
1218
1219
1220
1221
1222
1223
1224
	.llseek		= generic_file_llseek,
	.read		= fuse_direct_read,
	.write		= fuse_direct_write,
	.open		= fuse_open,
	.flush		= fuse_flush,
	.release	= fuse_release,
	.fsync		= fuse_fsync,
1225
	.lock		= fuse_file_lock,
1226
	.flock		= fuse_file_flock,
1227
	/* no mmap and splice_read */
Miklos Szeredi's avatar
Miklos Szeredi committed
1228
1229
};

1230
static const struct address_space_operations fuse_file_aops  = {
1231
	.readpage	= fuse_readpage,
Miklos Szeredi's avatar
Miklos Szeredi committed
1232
1233
	.writepage	= fuse_writepage,
	.launder_page	= fuse_launder_page,
Nick Piggin's avatar
Nick Piggin committed
1234
1235
	.write_begin	= fuse_write_begin,
	.write_end	= fuse_write_end,
1236
	.readpages	= fuse_readpages,
Miklos Szeredi's avatar
Miklos Szeredi committed
1237
	.set_page_dirty	= __set_page_dirty_nobuffers,
1238
	.bmap		= fuse_bmap,
1239
1240
1241
1242
};

void fuse_init_file_inode(struct inode *inode)
{
1243
1244
	inode->i_fop = &fuse_file_operations;
	inode->i_data.a_ops = &fuse_file_aops;
1245
}