file.c 15.2 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
14

  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>

15
static const struct file_operations fuse_direct_io_file_operations;
16

17
18
static int fuse_send_open(struct inode *inode, struct file *file, int isdir,
			  struct fuse_open_out *outargp)
19
20
21
{
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_open_in inarg;
22
23
24
	struct fuse_req *req;
	int err;

25
26
27
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78

	memset(&inarg, 0, sizeof(inarg));
	inarg.flags = file->f_flags & ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
	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) {
		ff->release_req = fuse_request_alloc();
		if (!ff->release_req) {
			kfree(ff);
			ff = NULL;
		}
	}
	return ff;
}

void fuse_file_free(struct fuse_file *ff)
{
	fuse_request_free(ff->release_req);
	kfree(ff);
}

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))
		invalidate_inode_pages(inode->i_mapping);
	ff->fh = outarg->fh;
	file->private_data = ff;
}

int fuse_open_common(struct inode *inode, struct file *file, int isdir)
{
79
80
81
82
	struct fuse_open_out outarg;
	struct fuse_file *ff;
	int err;

Miklos Szeredi's avatar
Miklos Szeredi committed
83
84
85
86
	/* VFS checks this, but only _after_ ->open() */
	if (file->f_flags & O_DIRECT)
		return -EINVAL;

87
88
89
90
91
92
93
	err = generic_file_open(inode, file);
	if (err)
		return err;

	/* If opening the root node, no lookup has been performed on
	   it, so the attributes must be refreshed */
	if (get_node_id(inode) == FUSE_ROOT_ID) {
94
		err = fuse_do_getattr(inode);
95
96
97
98
		if (err)
		 	return err;
	}

99
	ff = fuse_file_alloc();
100
	if (!ff)
101
		return -ENOMEM;
102

103
104
105
106
107
108
109
	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);
110
111
112
113
114
	}

	return err;
}

115
116
struct fuse_req *fuse_release_fill(struct fuse_file *ff, u64 nodeid, int flags,
				   int opcode)
117
{
118
	struct fuse_req *req = ff->release_req;
119
120
121
	struct fuse_release_in *inarg = &req->misc.release_in;

	inarg->fh = ff->fh;
122
	inarg->flags = flags;
123
	req->in.h.opcode = opcode;
124
	req->in.h.nodeid = nodeid;
125
126
127
128
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(struct fuse_release_in);
	req->in.args[0].value = inarg;
	kfree(ff);
129
130

	return req;
131
132
133
134
135
136
137
}

int fuse_release_common(struct inode *inode, struct file *file, int isdir)
{
	struct fuse_file *ff = file->private_data;
	if (ff) {
		struct fuse_conn *fc = get_fuse_conn(inode);
138
139
140
141
142
143
144
145
146
		struct fuse_req *req;

		req = fuse_release_fill(ff, get_node_id(inode), file->f_flags,
					isdir ? FUSE_RELEASEDIR : FUSE_RELEASE);

		/* Hold vfsmount and dentry until release is finished */
		req->vfsmount = mntget(file->f_vfsmnt);
		req->dentry = dget(file->f_dentry);
		request_send_background(fc, req);
147
	}
148
149
150
151
152

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

153
154
155
156
157
158
159
160
161
162
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);
}

163
static int fuse_flush(struct file *file, fl_owner_t id)
164
165
166
167
168
169
170
171
{
	struct inode *inode = file->f_dentry->d_inode;
	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;

172
173
174
	if (is_bad_inode(inode))
		return -EIO;

175
176
177
	if (fc->no_flush)
		return 0;

178
179
180
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
181
182
183
184
185
186
187
188

	memset(&inarg, 0, sizeof(inarg));
	inarg.fh = ff->fh;
	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;
189
	request_send(fc, req);
190
191
192
193
194
195
196
197
198
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS) {
		fc->no_flush = 1;
		err = 0;
	}
	return err;
}

199
200
int fuse_fsync_common(struct file *file, struct dentry *de, int datasync,
		      int isdir)
201
202
203
204
205
206
207
208
{
	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;

209
210
211
	if (is_bad_inode(inode))
		return -EIO;

212
	if ((!isdir && fc->no_fsync) || (isdir && fc->no_fsyncdir))
213
214
		return 0;

215
216
217
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
218
219
220
221

	memset(&inarg, 0, sizeof(inarg));
	inarg.fh = ff->fh;
	inarg.fsync_flags = datasync ? 1 : 0;
222
	req->in.h.opcode = isdir ? FUSE_FSYNCDIR : FUSE_FSYNC;
223
224
225
226
227
228
229
230
	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) {
231
232
233
234
		if (isdir)
			fc->no_fsyncdir = 1;
		else
			fc->no_fsync = 1;
235
236
237
238
239
		err = 0;
	}
	return err;
}

240
241
242
243
244
static int fuse_fsync(struct file *file, struct dentry *de, int datasync)
{
	return fuse_fsync_common(file, de, datasync, 0);
}

245
246
void fuse_read_fill(struct fuse_req *req, struct file *file,
		    struct inode *inode, loff_t pos, size_t count, int opcode)
247
248
{
	struct fuse_file *ff = file->private_data;
249
	struct fuse_read_in *inarg = &req->misc.read_in;
250

251
252
253
254
	inarg->fh = ff->fh;
	inarg->offset = pos;
	inarg->size = count;
	req->in.h.opcode = opcode;
255
256
257
	req->in.h.nodeid = get_node_id(inode);
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(struct fuse_read_in);
258
	req->in.args[0].value = inarg;
259
260
261
262
263
264
	req->out.argpages = 1;
	req->out.argvar = 1;
	req->out.numargs = 1;
	req->out.args[0].size = count;
}

265
266
static size_t fuse_send_read(struct fuse_req *req, struct file *file,
			     struct inode *inode, loff_t pos, size_t count)
267
{
268
269
270
271
	struct fuse_conn *fc = get_fuse_conn(inode);
	fuse_read_fill(req, file, inode, pos, count, FUSE_READ);
	request_send(fc, req);
	return req->out.args[0].size;
272
273
}

274
275
276
277
static int fuse_readpage(struct file *file, struct page *page)
{
	struct inode *inode = page->mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
278
279
280
281
282
283
284
	struct fuse_req *req;
	int err;

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

285
286
287
	req = fuse_get_req(fc);
	err = PTR_ERR(req);
	if (IS_ERR(req))
288
289
290
291
292
		goto out;

	req->out.page_zeroing = 1;
	req->num_pages = 1;
	req->pages[0] = page;
293
	fuse_send_read(req, file, inode, page_offset(page), PAGE_CACHE_SIZE);
294
295
296
297
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err)
		SetPageUptodate(page);
298
	fuse_invalidate_attr(inode); /* atime changed */
299
300
301
302
303
 out:
	unlock_page(page);
	return err;
}

304
static void fuse_readpages_end(struct fuse_conn *fc, struct fuse_req *req)
305
{
306
307
308
309
	int i;

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

310
311
312
313
	for (i = 0; i < req->num_pages; i++) {
		struct page *page = req->pages[i];
		if (!req->out.h.error)
			SetPageUptodate(page);
314
315
		else
			SetPageError(page);
316
317
		unlock_page(page);
	}
318
319
320
321
322
323
324
325
326
327
328
	fuse_put_request(fc, req);
}

static void fuse_send_readpages(struct fuse_req *req, struct file *file,
				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;
	fuse_read_fill(req, file, inode, pos, count, FUSE_READ);
329
	if (fc->async_read) {
330
331
		get_file(file);
		req->file = file;
332
333
334
335
336
337
		req->end = fuse_readpages_end;
		request_send_background(fc, req);
	} else {
		request_send(fc, req);
		fuse_readpages_end(fc, req);
	}
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
}

struct fuse_readpages_data {
	struct fuse_req *req;
	struct file *file;
	struct inode *inode;
};

static int fuse_readpages_fill(void *_data, struct page *page)
{
	struct fuse_readpages_data *data = _data;
	struct fuse_req *req = data->req;
	struct inode *inode = data->inode;
	struct fuse_conn *fc = get_fuse_conn(inode);

	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)) {
357
		fuse_send_readpages(req, data->file, inode);
358
359
		data->req = req = fuse_get_req(fc);
		if (IS_ERR(req)) {
360
			unlock_page(page);
361
			return PTR_ERR(req);
362
363
364
365
366
367
368
369
370
371
372
373
374
375
		}
	}
	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);
	struct fuse_readpages_data data;
	int err;
376
377
378
379

	if (is_bad_inode(inode))
		return -EIO;

380
381
	data.file = file;
	data.inode = inode;
382
383
384
	data.req = fuse_get_req(fc);
	if (IS_ERR(data.req))
		return PTR_ERR(data.req);
385
386

	err = read_cache_pages(mapping, pages, fuse_readpages_fill, &data);
387
388
389
390
391
392
	if (!err) {
		if (data.req->num_pages)
			fuse_send_readpages(data.req, file, inode);
		else
			fuse_put_request(fc, data.req);
	}
393
394
395
	return err;
}

396
397
static size_t fuse_send_write(struct fuse_req *req, struct file *file,
			      struct inode *inode, loff_t pos, size_t count)
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
{
	struct fuse_conn *fc = get_fuse_conn(inode);
	struct fuse_file *ff = file->private_data;
	struct fuse_write_in inarg;
	struct fuse_write_out outarg;

	memset(&inarg, 0, sizeof(struct fuse_write_in));
	inarg.fh = ff->fh;
	inarg.offset = pos;
	inarg.size = count;
	req->in.h.opcode = FUSE_WRITE;
	req->in.h.nodeid = get_node_id(inode);
	req->in.argpages = 1;
	req->in.numargs = 2;
	req->in.args[0].size = sizeof(struct fuse_write_in);
	req->in.args[0].value = &inarg;
	req->in.args[1].size = count;
	req->out.numargs = 1;
	req->out.args[0].size = sizeof(struct fuse_write_out);
	req->out.args[0].value = &outarg;
418
	request_send(fc, req);
419
420
421
422
423
424
425
426
427
428
429
430
431
432
	return outarg.size;
}

static int fuse_prepare_write(struct file *file, struct page *page,
			      unsigned offset, unsigned to)
{
	/* No op */
	return 0;
}

static int fuse_commit_write(struct file *file, struct page *page,
			     unsigned offset, unsigned to)
{
	int err;
433
	size_t nres;
434
435
436
	unsigned count = to - offset;
	struct inode *inode = page->mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
437
	loff_t pos = page_offset(page) + offset;
438
439
440
441
442
	struct fuse_req *req;

	if (is_bad_inode(inode))
		return -EIO;

443
444
445
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463

	req->num_pages = 1;
	req->pages[0] = page;
	req->page_offset = offset;
	nres = fuse_send_write(req, file, inode, pos, count);
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err && nres != count)
		err = -EIO;
	if (!err) {
		pos += count;
		if (pos > i_size_read(inode))
			i_size_write(inode, pos);

		if (offset == 0 && to == PAGE_CACHE_SIZE) {
			clear_page_dirty(page);
			SetPageUptodate(page);
		}
464
465
	}
	fuse_invalidate_attr(inode);
466
467
468
	return err;
}

Miklos Szeredi's avatar
Miklos Szeredi committed
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
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;
494
	npages = min(max(npages, 1), FUSE_MAX_PAGES_PER_REQ);
Miklos Szeredi's avatar
Miklos Szeredi committed
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
	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)
{
	struct inode *inode = file->f_dentry->d_inode;
	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;
515
516
517
518
519
	struct fuse_req *req;

	if (is_bad_inode(inode))
		return -EIO;

520
521
522
	req = fuse_get_req(fc);
	if (IS_ERR(req))
		return PTR_ERR(req);
Miklos Szeredi's avatar
Miklos Szeredi committed
523
524
525
526
527
528
529
530
531

	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;
		}
532
533
		nbytes = (req->num_pages << PAGE_SHIFT) - req->page_offset;
		nbytes = min(count, nbytes);
Miklos Szeredi's avatar
Miklos Szeredi committed
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
		if (write)
			nres = fuse_send_write(req, file, inode, pos, nbytes);
		else
			nres = fuse_send_read(req, file, inode, pos, nbytes);
		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;
553
554
555
556
557
558
		if (count) {
			fuse_put_request(fc, req);
			req = fuse_get_req(fc);
			if (IS_ERR(req))
				break;
		}
Miklos Szeredi's avatar
Miklos Szeredi committed
559
560
561
562
563
564
	}
	fuse_put_request(fc, req);
	if (res > 0) {
		if (write && pos > i_size_read(inode))
			i_size_write(inode, pos);
		*ppos = pos;
565
566
	}
	fuse_invalidate_attr(inode);
Miklos Szeredi's avatar
Miklos Szeredi committed
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582

	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)
{
	struct inode *inode = file->f_dentry->d_inode;
	ssize_t res;
	/* Don't allow parallel writes to the same file */
583
	mutex_lock(&inode->i_mutex);
Miklos Szeredi's avatar
Miklos Szeredi committed
584
	res = fuse_direct_io(file, buf, count, ppos, 1);
585
	mutex_unlock(&inode->i_mutex);
Miklos Szeredi's avatar
Miklos Szeredi committed
586
587
588
	return res;
}

589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
static int fuse_file_mmap(struct file *file, struct vm_area_struct *vma)
{
	if ((vma->vm_flags & VM_SHARED)) {
		if ((vma->vm_flags & VM_WRITE))
			return -ENODEV;
		else
			vma->vm_flags &= ~VM_MAYWRITE;
	}
	return generic_file_mmap(file, vma);
}

static int fuse_set_page_dirty(struct page *page)
{
	printk("fuse_set_page_dirty: should not happen\n");
	dump_stack();
	return 0;
}

607
static const struct file_operations fuse_file_operations = {
608
609
610
611
612
613
614
615
616
617
618
	.llseek		= generic_file_llseek,
	.read		= generic_file_read,
	.write		= generic_file_write,
	.mmap		= fuse_file_mmap,
	.open		= fuse_open,
	.flush		= fuse_flush,
	.release	= fuse_release,
	.fsync		= fuse_fsync,
	.sendfile	= generic_file_sendfile,
};

619
static const struct file_operations fuse_direct_io_file_operations = {
Miklos Szeredi's avatar
Miklos Szeredi committed
620
621
622
623
624
625
626
627
628
629
	.llseek		= generic_file_llseek,
	.read		= fuse_direct_read,
	.write		= fuse_direct_write,
	.open		= fuse_open,
	.flush		= fuse_flush,
	.release	= fuse_release,
	.fsync		= fuse_fsync,
	/* no mmap and sendfile */
};

630
631
632
633
static struct address_space_operations fuse_file_aops  = {
	.readpage	= fuse_readpage,
	.prepare_write	= fuse_prepare_write,
	.commit_write	= fuse_commit_write,
634
	.readpages	= fuse_readpages,
635
636
637
638
639
	.set_page_dirty	= fuse_set_page_dirty,
};

void fuse_init_file_inode(struct inode *inode)
{
640
641
	inode->i_fop = &fuse_file_operations;
	inode->i_data.a_ops = &fuse_file_aops;
642
}