file.c 15.1 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
/*
  FUSE: Filesystem in Userspace
  Copyright (C) 2001-2005  Miklos Szeredi <miklos@szeredi.hu>

  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
16
static struct file_operations fuse_direct_io_file_operations;

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
25
26
27
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
79
	struct fuse_req *req;
	int err;

	req = fuse_get_request(fc);
	if (!req)
		return -EINTR;

	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->inode = 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)
{
80
81
82
83
	struct fuse_open_out outarg;
	struct fuse_file *ff;
	int err;

Miklos Szeredi's avatar
Miklos Szeredi committed
84
85
86
87
	/* VFS checks this, but only _after_ ->open() */
	if (file->f_flags & O_DIRECT)
		return -EINVAL;

88
89
90
91
92
93
94
	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) {
95
		err = fuse_do_getattr(inode);
96
97
98
99
		if (err)
		 	return err;
	}

100
	ff = fuse_file_alloc();
101
	if (!ff)
102
		return -ENOMEM;
103

104
105
106
107
108
109
110
	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);
111
112
113
114
115
	}

	return err;
}

116
117
118
119
120
121
122
123
/* Special case for failed iget in CREATE */
static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req)
{
	u64 nodeid = req->in.h.nodeid;
	fuse_reset_request(req);
	fuse_send_forget(fc, req, nodeid, 1);
}

124
125
void fuse_send_release(struct fuse_conn *fc, struct fuse_file *ff,
		       u64 nodeid, struct inode *inode, int flags, int isdir)
126
{
127
	struct fuse_req * req = ff->release_req;
128
129
130
	struct fuse_release_in *inarg = &req->misc.release_in;

	inarg->fh = ff->fh;
131
	inarg->flags = flags;
132
	req->in.h.opcode = isdir ? FUSE_RELEASEDIR : FUSE_RELEASE;
133
	req->in.h.nodeid = nodeid;
134
135
136
137
138
	req->inode = inode;
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(struct fuse_release_in);
	req->in.args[0].value = inarg;
	request_send_background(fc, req);
139
140
	if (!inode)
		req->end = fuse_release_end;
141
	kfree(ff);
142
143
144
145
146
147
148
149
150
151
}

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);
		u64 nodeid = get_node_id(inode);
		fuse_send_release(fc, ff, nodeid, inode, file->f_flags, isdir);
	}
152
153
154
155
156

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

157
158
159
160
161
162
163
164
165
166
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);
}

167
168
169
170
171
172
173
174
175
static int fuse_flush(struct file *file)
{
	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;

176
177
178
	if (is_bad_inode(inode))
		return -EIO;

179
180
181
	if (fc->no_flush)
		return 0;

182
	req = fuse_get_request(fc);
183
184
185
186
187
188
189
190
191
192
193
194
	if (!req)
		return -EINTR;

	memset(&inarg, 0, sizeof(inarg));
	inarg.fh = ff->fh;
	req->in.h.opcode = FUSE_FLUSH;
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->file = file;
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(inarg);
	req->in.args[0].value = &inarg;
195
	request_send(fc, req);
196
197
198
199
200
201
202
203
204
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (err == -ENOSYS) {
		fc->no_flush = 1;
		err = 0;
	}
	return err;
}

205
206
int fuse_fsync_common(struct file *file, struct dentry *de, int datasync,
		      int isdir)
207
208
209
210
211
212
213
214
{
	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;

215
216
217
	if (is_bad_inode(inode))
		return -EIO;

218
	if ((!isdir && fc->no_fsync) || (isdir && fc->no_fsyncdir))
219
220
221
222
		return 0;

	req = fuse_get_request(fc);
	if (!req)
223
		return -EINTR;
224
225
226
227

	memset(&inarg, 0, sizeof(inarg));
	inarg.fh = ff->fh;
	inarg.fsync_flags = datasync ? 1 : 0;
228
	req->in.h.opcode = isdir ? FUSE_FSYNCDIR : FUSE_FSYNC;
229
230
231
232
233
234
235
236
237
238
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->file = file;
	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) {
239
240
241
242
		if (isdir)
			fc->no_fsyncdir = 1;
		else
			fc->no_fsync = 1;
243
244
245
246
247
		err = 0;
	}
	return err;
}

248
249
250
251
252
static int fuse_fsync(struct file *file, struct dentry *de, int datasync)
{
	return fuse_fsync_common(file, de, datasync, 0);
}

253
254
void fuse_read_fill(struct fuse_req *req, struct file *file,
		    struct inode *inode, loff_t pos, size_t count, int opcode)
255
256
{
	struct fuse_file *ff = file->private_data;
257
	struct fuse_read_in *inarg = &req->misc.read_in;
258

259
260
261
262
	inarg->fh = ff->fh;
	inarg->offset = pos;
	inarg->size = count;
	req->in.h.opcode = opcode;
263
264
265
266
267
	req->in.h.nodeid = get_node_id(inode);
	req->inode = inode;
	req->file = file;
	req->in.numargs = 1;
	req->in.args[0].size = sizeof(struct fuse_read_in);
268
	req->in.args[0].value = inarg;
269
270
271
272
273
274
	req->out.argpages = 1;
	req->out.argvar = 1;
	req->out.numargs = 1;
	req->out.args[0].size = count;
}

275
276
static size_t fuse_send_read(struct fuse_req *req, struct file *file,
			     struct inode *inode, loff_t pos, size_t count)
277
{
278
279
280
281
	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;
282
283
}

284
285
286
287
static int fuse_readpage(struct file *file, struct page *page)
{
	struct inode *inode = page->mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
288
289
290
291
292
293
294
295
296
	struct fuse_req *req;
	int err;

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

	err = -EINTR;
	req = fuse_get_request(fc);
297
298
299
300
301
302
	if (!req)
		goto out;

	req->out.page_zeroing = 1;
	req->num_pages = 1;
	req->pages[0] = page;
303
	fuse_send_read(req, file, inode, page_offset(page), PAGE_CACHE_SIZE);
304
305
306
307
	err = req->out.h.error;
	fuse_put_request(fc, req);
	if (!err)
		SetPageUptodate(page);
308
	fuse_invalidate_attr(inode); /* atime changed */
309
310
311
312
313
 out:
	unlock_page(page);
	return err;
}

314
static void fuse_readpages_end(struct fuse_conn *fc, struct fuse_req *req)
315
{
316
317
318
319
	int i;

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

320
321
322
323
	for (i = 0; i < req->num_pages; i++) {
		struct page *page = req->pages[i];
		if (!req->out.h.error)
			SetPageUptodate(page);
324
325
		else
			SetPageError(page);
326
327
		unlock_page(page);
	}
328
329
330
331
332
333
334
335
336
337
338
339
340
	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;
	req->end = fuse_readpages_end;
	fuse_read_fill(req, file, inode, pos, count, FUSE_READ);
	request_send_background(fc, req);
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
}

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)) {
360
361
362
		fuse_send_readpages(req, data->file, inode);
		data->req = req = fuse_get_request(fc);
		if (!req) {
363
			unlock_page(page);
364
			return -EINTR;
365
366
367
368
369
370
371
372
373
374
375
376
377
378
		}
	}
	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;
379
380
381
382

	if (is_bad_inode(inode))
		return -EIO;

383
384
	data.file = file;
	data.inode = inode;
385
	data.req = fuse_get_request(fc);
386
387
388
389
	if (!data.req)
		return -EINTR;

	err = read_cache_pages(mapping, pages, fuse_readpages_fill, &data);
390
391
	if (!err)
		fuse_send_readpages(data.req, file, inode);
392
393
394
	return err;
}

395
396
static size_t fuse_send_write(struct fuse_req *req, struct file *file,
			      struct inode *inode, loff_t pos, size_t count)
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
{
	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->inode = inode;
	req->file = file;
	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;
419
	request_send(fc, req);
420
421
422
423
424
425
426
427
428
429
430
431
432
433
	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;
434
	size_t nres;
435
436
437
	unsigned count = to - offset;
	struct inode *inode = page->mapping->host;
	struct fuse_conn *fc = get_fuse_conn(inode);
438
	loff_t pos = page_offset(page) + offset;
439
440
441
442
443
444
	struct fuse_req *req;

	if (is_bad_inode(inode))
		return -EIO;

	req = fuse_get_request(fc);
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
	if (!req)
		return -EINTR;

	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);
		}
465
466
	}
	fuse_invalidate_attr(inode);
467
468
469
	return err;
}

Miklos Szeredi's avatar
Miklos Szeredi committed
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
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;
495
	npages = min(max(npages, 1), FUSE_MAX_PAGES_PER_REQ);
Miklos Szeredi's avatar
Miklos Szeredi committed
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
	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;
516
517
518
519
520
521
	struct fuse_req *req;

	if (is_bad_inode(inode))
		return -EIO;

	req = fuse_get_request(fc);
Miklos Szeredi's avatar
Miklos Szeredi committed
522
	if (!req)
523
		return -EINTR;
Miklos Szeredi's avatar
Miklos Szeredi committed
524
525
526
527
528
529
530
531
532

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

	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 */
580
	mutex_lock(&inode->i_mutex);
Miklos Szeredi's avatar
Miklos Szeredi committed
581
	res = fuse_direct_io(file, buf, count, ppos, 1);
582
	mutex_unlock(&inode->i_mutex);
Miklos Szeredi's avatar
Miklos Szeredi committed
583
584
585
	return res;
}

586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
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;
}

static struct file_operations fuse_file_operations = {
	.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,
};

Miklos Szeredi's avatar
Miklos Szeredi committed
616
617
618
619
620
621
622
623
624
625
626
static struct file_operations fuse_direct_io_file_operations = {
	.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 */
};

627
628
629
630
static struct address_space_operations fuse_file_aops  = {
	.readpage	= fuse_readpage,
	.prepare_write	= fuse_prepare_write,
	.commit_write	= fuse_commit_write,
631
	.readpages	= fuse_readpages,
632
633
634
635
636
	.set_page_dirty	= fuse_set_page_dirty,
};

void fuse_init_file_inode(struct inode *inode)
{
637
638
	inode->i_fop = &fuse_file_operations;
	inode->i_data.a_ops = &fuse_file_aops;
639
}