Subversion Repositories HelenOS

Rev

Rev 3119 | Rev 3257 | Go to most recent revision | Details | Compare with Previous | Last modification | View Log | RSS feed

Rev Author Line No. Line
2627 jermar 1
/*
2793 jermar 2
 * Copyright (c) 2008 Jakub Jermar
2627 jermar 3
 * All rights reserved.
4
 *
5
 * Redistribution and use in source and binary forms, with or without
6
 * modification, are permitted provided that the following conditions
7
 * are met:
8
 *
9
 * - Redistributions of source code must retain the above copyright
10
 *   notice, this list of conditions and the following disclaimer.
11
 * - Redistributions in binary form must reproduce the above copyright
12
 *   notice, this list of conditions and the following disclaimer in the
13
 *   documentation and/or other materials provided with the distribution.
14
 * - The name of the author may not be used to endorse or promote products
15
 *   derived from this software without specific prior written permission.
16
 *
17
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18
 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
19
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
20
 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
21
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
22
 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
26
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
27
 */
28
 
29
/** @addtogroup fs
30
 * @{
31
 */ 
32
 
33
/**
34
 * @file	fat_ops.c
35
 * @brief	Implementation of VFS operations for the FAT file system server.
36
 */
37
 
38
#include "fat.h"
2638 jermar 39
#include "../../vfs/vfs.h"
2793 jermar 40
#include <libfs.h>
2627 jermar 41
#include <ipc/ipc.h>
42
#include <async.h>
43
#include <errno.h>
2793 jermar 44
#include <string.h>
2798 jermar 45
#include <byteorder.h>
2831 jermar 46
#include <libadt/hash_table.h>
47
#include <libadt/list.h>
48
#include <assert.h>
2856 jermar 49
#include <futex.h>
2627 jermar 50
 
2843 jermar 51
#define BS_BLOCK		0
3253 jermar 52
#define BS_SIZE			512
2843 jermar 53
 
2951 jermar 54
/** Futex protecting the list of cached free FAT nodes. */
55
static futex_t ffn_futex = FUTEX_INITIALIZER;
2843 jermar 56
 
2951 jermar 57
/** List of cached free FAT nodes. */
58
static LIST_INITIALIZE(ffn_head);
59
 
2638 jermar 60
#define FAT_NAME_LEN		8
61
#define FAT_EXT_LEN		3
62
 
63
#define FAT_PAD			' ' 
64
 
65
#define FAT_DENTRY_UNUSED	0x00
66
#define FAT_DENTRY_E5_ESC	0x05
67
#define FAT_DENTRY_DOT		0x2e
68
#define FAT_DENTRY_ERASED	0xe5
69
 
2793 jermar 70
static void dentry_name_canonify(fat_dentry_t *d, char *buf)
2638 jermar 71
{
2793 jermar 72
	int i;
2639 jermar 73
 
2793 jermar 74
	for (i = 0; i < FAT_NAME_LEN; i++) {
75
		if (d->name[i] == FAT_PAD) {
76
			buf++;
77
			break;
2639 jermar 78
		}
2793 jermar 79
		if (d->name[i] == FAT_DENTRY_E5_ESC)
80
			*buf++ = 0xe5;
81
		else
82
			*buf++ = d->name[i];
83
	}
84
	if (d->ext[0] != FAT_PAD)
85
		*buf++ = '.';
86
	for (i = 0; i < FAT_EXT_LEN; i++) {
87
		if (d->ext[i] == FAT_PAD) {
88
			*buf = '\0';
89
			return;
90
		}
91
		if (d->ext[i] == FAT_DENTRY_E5_ESC)
92
			*buf++ = 0xe5;
93
		else
94
			*buf++ = d->ext[i];
95
	}
96
}
97
 
2910 jermar 98
/* TODO move somewhere else */
2822 jermar 99
typedef struct {
100
	void *data;
101
} block_t;
102
 
3253 jermar 103
static block_t *block_get(dev_handle_t dev_handle, off_t offset, size_t bs)
2793 jermar 104
{
105
	return NULL;	/* TODO */
106
}
107
 
2822 jermar 108
static void block_put(block_t *block)
2793 jermar 109
{
110
	/* TODO */
111
}
112
 
2859 jermar 113
#define FAT_BS(b)		((fat_bs_t *)((b)->data))
114
 
2864 jermar 115
#define FAT_CLST_RES0	0x0000
3119 jermar 116
#define FAT_CLST_RES1	0x0001
2859 jermar 117
#define FAT_CLST_FIRST	0x0002
118
#define FAT_CLST_BAD	0xfff7
119
#define FAT_CLST_LAST1	0xfff8
120
#define FAT_CLST_LAST8  0xffff
121
 
3119 jermar 122
/* internally used to mark root directory's parent */
123
#define FAT_CLST_ROOTPAR	FAT_CLST_RES0
124
/* internally used to mark root directory */
125
#define FAT_CLST_ROOT		FAT_CLST_RES1
126
 
2891 jermar 127
#define fat_block_get(np, off) \
128
    _fat_block_get((np)->idx->dev_handle, (np)->firstc, (off))
129
 
130
static block_t *
2893 jermar 131
_fat_block_get(dev_handle_t dev_handle, fat_cluster_t firstc, off_t offset)
2859 jermar 132
{
133
	block_t *bb;
134
	block_t *b;
135
	unsigned bps;
136
	unsigned spc;
137
	unsigned rscnt;		/* block address of the first FAT */
138
	unsigned fatcnt;
139
	unsigned rde;
140
	unsigned rds;		/* root directory size */
141
	unsigned sf;
142
	unsigned ssa;		/* size of the system area */
143
	unsigned clusters;
2893 jermar 144
	fat_cluster_t clst = firstc;
2859 jermar 145
	unsigned i;
146
 
3253 jermar 147
	bb = block_get(dev_handle, BS_BLOCK, BS_SIZE);
2859 jermar 148
	bps = uint16_t_le2host(FAT_BS(bb)->bps);
149
	spc = FAT_BS(bb)->spc;
150
	rscnt = uint16_t_le2host(FAT_BS(bb)->rscnt);
151
	fatcnt = FAT_BS(bb)->fatcnt;
152
	rde = uint16_t_le2host(FAT_BS(bb)->root_ent_max);
153
	sf = uint16_t_le2host(FAT_BS(bb)->sec_per_fat);
154
	block_put(bb);
155
 
156
	rds = (sizeof(fat_dentry_t) * rde) / bps;
157
	rds += ((sizeof(fat_dentry_t) * rde) % bps != 0);
158
	ssa = rscnt + fatcnt * sf + rds;
159
 
3119 jermar 160
	if (firstc == FAT_CLST_ROOT) {
2859 jermar 161
		/* root directory special case */
162
		assert(offset < rds);
3253 jermar 163
		b = block_get(dev_handle, rscnt + fatcnt * sf + offset, bps);
2859 jermar 164
		return b;
165
	}
166
 
167
	clusters = offset / spc; 
168
	for (i = 0; i < clusters; i++) {
169
		unsigned fsec;	/* sector offset relative to FAT1 */
170
		unsigned fidx;	/* FAT1 entry index */
171
 
172
		assert(clst >= FAT_CLST_FIRST && clst < FAT_CLST_BAD);
2864 jermar 173
		fsec = (clst * sizeof(fat_cluster_t)) / bps;
174
		fidx = clst % (bps / sizeof(fat_cluster_t));
2859 jermar 175
		/* read FAT1 */
3253 jermar 176
		b = block_get(dev_handle, rscnt + fsec, bps);
2864 jermar 177
		clst = uint16_t_le2host(((fat_cluster_t *)b->data)[fidx]);
2859 jermar 178
		assert(clst != FAT_CLST_BAD);
179
		assert(clst < FAT_CLST_LAST1);
180
		block_put(b);
181
	}
182
 
2891 jermar 183
	b = block_get(dev_handle, ssa + (clst - FAT_CLST_FIRST) * spc +
3253 jermar 184
	    offset % spc, bps);
2859 jermar 185
 
186
	return b;
187
}
188
 
2831 jermar 189
static void fat_node_initialize(fat_node_t *node)
2793 jermar 190
{
2951 jermar 191
	futex_initialize(&node->lock, 1);
2864 jermar 192
	node->idx = NULL;
2831 jermar 193
	node->type = 0;
194
	link_initialize(&node->ffn_link);
195
	node->size = 0;
196
	node->lnkcnt = 0;
197
	node->refcnt = 0;
198
	node->dirty = false;
2793 jermar 199
}
200
 
2843 jermar 201
static uint16_t fat_bps_get(dev_handle_t dev_handle)
202
{
203
	block_t *bb;
204
	uint16_t bps;
205
 
3253 jermar 206
	bb = block_get(dev_handle, BS_BLOCK, BS_SIZE);
2843 jermar 207
	assert(bb != NULL);
2859 jermar 208
	bps = uint16_t_le2host(FAT_BS(bb)->bps);
2843 jermar 209
	block_put(bb);
210
 
211
	return bps;
212
}
213
 
2845 jermar 214
typedef enum {
215
	FAT_DENTRY_SKIP,
216
	FAT_DENTRY_LAST,
217
	FAT_DENTRY_VALID
218
} fat_dentry_clsf_t;
219
 
220
static fat_dentry_clsf_t fat_classify_dentry(fat_dentry_t *d)
221
{
222
	if (d->attr & FAT_ATTR_VOLLABEL) {
223
		/* volume label entry */
224
		return FAT_DENTRY_SKIP;
225
	}
226
	if (d->name[0] == FAT_DENTRY_ERASED) {
227
		/* not-currently-used entry */
228
		return FAT_DENTRY_SKIP;
229
	}
230
	if (d->name[0] == FAT_DENTRY_UNUSED) {
231
		/* never used entry */
232
		return FAT_DENTRY_LAST;
233
	}
234
	if (d->name[0] == FAT_DENTRY_DOT) {
235
		/*
236
		 * Most likely '.' or '..'.
237
		 * It cannot occur in a regular file name.
238
		 */
239
		return FAT_DENTRY_SKIP;
240
	}
241
	return FAT_DENTRY_VALID;
242
}
243
 
2893 jermar 244
static void fat_node_sync(fat_node_t *node)
2831 jermar 245
{
246
	/* TODO */
247
}
248
 
2951 jermar 249
/** Internal version of fat_node_get().
250
 *
251
 * @param idxp		Locked index structure.
252
 */
253
static void *fat_node_get_core(fat_idx_t *idxp)
2831 jermar 254
{
2891 jermar 255
	block_t *b;
256
	fat_dentry_t *d;
257
	fat_node_t *nodep;
258
	unsigned bps;
259
	unsigned dps;
260
 
2951 jermar 261
	if (idxp->nodep) {
2891 jermar 262
		/*
263
		 * We are lucky.
264
		 * The node is already instantiated in memory.
265
		 */
2951 jermar 266
		futex_down(&idxp->nodep->lock);
267
		if (!idxp->nodep->refcnt++)
2910 jermar 268
			list_remove(&nodep->ffn_link);
2951 jermar 269
		futex_up(&idxp->nodep->lock);
270
		return idxp->nodep;
2891 jermar 271
	}
272
 
273
	/*
274
	 * We must instantiate the node from the file system.
275
	 */
276
 
2951 jermar 277
	assert(idxp->pfc);
2891 jermar 278
 
2951 jermar 279
	futex_down(&ffn_futex);
2893 jermar 280
	if (!list_empty(&ffn_head)) {
2951 jermar 281
		/* Try to use a cached free node structure. */
282
		fat_idx_t *idxp_tmp;
2893 jermar 283
		nodep = list_get_instance(ffn_head.next, fat_node_t, ffn_link);
2951 jermar 284
		if (futex_trydown(&nodep->lock) == ESYNCH_WOULD_BLOCK)
285
			goto skip_cache;
286
		idxp_tmp = nodep->idx;
287
		if (futex_trydown(&idxp_tmp->lock) == ESYNCH_WOULD_BLOCK) {
288
			futex_up(&nodep->lock);
289
			goto skip_cache;
290
		}
291
		list_remove(&nodep->ffn_link);
292
		futex_up(&ffn_futex);
2893 jermar 293
		if (nodep->dirty)
294
			fat_node_sync(nodep);
2951 jermar 295
		idxp_tmp->nodep = NULL;
296
		futex_up(&nodep->lock);
297
		futex_up(&idxp_tmp->lock);
2893 jermar 298
	} else {
2951 jermar 299
skip_cache:
2893 jermar 300
		/* Try to allocate a new node structure. */
2951 jermar 301
		futex_up(&ffn_futex);
2893 jermar 302
		nodep = (fat_node_t *)malloc(sizeof(fat_node_t));
303
		if (!nodep)
304
			return NULL;
305
	}
2891 jermar 306
	fat_node_initialize(nodep);
307
 
2951 jermar 308
	bps = fat_bps_get(idxp->dev_handle);
2891 jermar 309
	dps = bps / sizeof(fat_dentry_t);
310
 
2893 jermar 311
	/* Read the block that contains the dentry of interest. */
2951 jermar 312
	b = _fat_block_get(idxp->dev_handle, idxp->pfc,
313
	    (idxp->pdi * sizeof(fat_dentry_t)) / bps);
2891 jermar 314
	assert(b);
315
 
2951 jermar 316
	d = ((fat_dentry_t *)b->data) + (idxp->pdi % dps);
2893 jermar 317
	if (d->attr & FAT_ATTR_SUBDIR) {
318
		/* 
319
		 * The only directory which does not have this bit set is the
320
		 * root directory itself. The root directory node is handled
321
		 * and initialized elsewhere.
322
		 */
323
		nodep->type = FAT_DIRECTORY;
324
	} else {
325
		nodep->type = FAT_FILE;
326
	}
327
	nodep->firstc = uint16_t_le2host(d->firstc);
328
	nodep->size = uint32_t_le2host(d->size);
329
	nodep->lnkcnt = 1;
330
	nodep->refcnt = 1;
331
 
332
	block_put(b);
333
 
334
	/* Link the idx structure with the node structure. */
2951 jermar 335
	nodep->idx = idxp;
336
	idxp->nodep = nodep;
2893 jermar 337
 
338
	return nodep;
2831 jermar 339
}
340
 
2951 jermar 341
/** Instantiate a FAT in-core node. */
342
static void *fat_node_get(dev_handle_t dev_handle, fs_index_t index)
343
{
344
	void *node;
345
	fat_idx_t *idxp;
346
 
347
	idxp = fat_idx_get_by_index(dev_handle, index);
348
	if (!idxp)
349
		return NULL;
350
	/* idxp->lock held */
351
	node = fat_node_get_core(idxp);
352
	futex_up(&idxp->lock);
353
	return node;
354
}
355
 
2852 jermar 356
static void fat_node_put(void *node)
357
{
2910 jermar 358
	fat_node_t *nodep = (fat_node_t *)node;
359
 
2951 jermar 360
	futex_down(&nodep->lock);
2910 jermar 361
	if (!--nodep->refcnt) {
2951 jermar 362
		futex_down(&ffn_futex);
2910 jermar 363
		list_append(&nodep->ffn_link, &ffn_head);
2951 jermar 364
		futex_up(&ffn_futex);
2910 jermar 365
	}
2951 jermar 366
	futex_up(&nodep->lock);
2852 jermar 367
}
368
 
2857 jermar 369
static void *fat_create(int flags)
370
{
371
	return NULL;	/* not supported at the moment */
372
}
373
 
2858 jermar 374
static int fat_destroy(void *node)
2857 jermar 375
{
2858 jermar 376
	return ENOTSUP;	/* not supported at the moment */
2857 jermar 377
}
378
 
379
static bool fat_link(void *prnt, void *chld, const char *name)
380
{
381
	return false;	/* not supported at the moment */
382
}
383
 
384
static int fat_unlink(void *prnt, void *chld)
385
{
386
	return ENOTSUP;	/* not supported at the moment */
387
}
388
 
2793 jermar 389
static void *fat_match(void *prnt, const char *component)
390
{
391
	fat_node_t *parentp = (fat_node_t *)prnt;
392
	char name[FAT_NAME_LEN + 1 + FAT_EXT_LEN + 1];
2822 jermar 393
	unsigned i, j;
2828 jermar 394
	unsigned bps;		/* bytes per sector */
2822 jermar 395
	unsigned dps;		/* dentries per sector */
396
	unsigned blocks;
2793 jermar 397
	fat_dentry_t *d;
2822 jermar 398
	block_t *b;
2793 jermar 399
 
2953 jermar 400
	futex_down(&parentp->idx->lock);
2864 jermar 401
	bps = fat_bps_get(parentp->idx->dev_handle);
2828 jermar 402
	dps = bps / sizeof(fat_dentry_t);
403
	blocks = parentp->size / bps + (parentp->size % bps != 0);
2822 jermar 404
	for (i = 0; i < blocks; i++) {
405
		unsigned dentries;
2793 jermar 406
 
2864 jermar 407
		b = fat_block_get(parentp, i);
2822 jermar 408
		dentries = (i == blocks - 1) ?
409
		    parentp->size % sizeof(fat_dentry_t) :
410
		    dps;
411
		for (j = 0; j < dentries; j++) { 
412
			d = ((fat_dentry_t *)b->data) + j;
2845 jermar 413
			switch (fat_classify_dentry(d)) {
414
			case FAT_DENTRY_SKIP:
2822 jermar 415
				continue;
2845 jermar 416
			case FAT_DENTRY_LAST:
2822 jermar 417
				block_put(b);
2953 jermar 418
				futex_up(&parentp->idx->lock);
2822 jermar 419
				return NULL;
2845 jermar 420
			default:
421
			case FAT_DENTRY_VALID:
422
				dentry_name_canonify(d, name);
423
				break;
2822 jermar 424
			}
425
			if (strcmp(name, component) == 0) {
426
				/* hit */
2951 jermar 427
				void *node;
2953 jermar 428
				/*
429
				 * Assume tree hierarchy for locking.  We
430
				 * already have the parent and now we are going
431
				 * to lock the child.  Never lock in the oposite
432
				 * order.
433
				 */
2890 jermar 434
				fat_idx_t *idx = fat_idx_get_by_pos(
2881 jermar 435
				    parentp->idx->dev_handle, parentp->firstc,
2864 jermar 436
				    i * dps + j);
2953 jermar 437
				futex_up(&parentp->idx->lock);
2890 jermar 438
				if (!idx) {
439
					/*
440
					 * Can happen if memory is low or if we
441
					 * run out of 32-bit indices.
442
					 */
443
					block_put(b);
444
					return NULL;
445
				}
2951 jermar 446
				node = fat_node_get_core(idx);
447
				futex_up(&idx->lock);
2822 jermar 448
				block_put(b);
449
				return node;
450
			}
2793 jermar 451
		}
2822 jermar 452
		block_put(b);
2639 jermar 453
	}
2953 jermar 454
	futex_up(&parentp->idx->lock);
2793 jermar 455
	return NULL;
2638 jermar 456
}
457
 
2831 jermar 458
static fs_index_t fat_index_get(void *node)
459
{
460
	fat_node_t *fnodep = (fat_node_t *)node;
461
	if (!fnodep)
462
		return 0;
2864 jermar 463
	return fnodep->idx->index;
2831 jermar 464
}
465
 
466
static size_t fat_size_get(void *node)
467
{
468
	return ((fat_node_t *)node)->size;
469
}
470
 
471
static unsigned fat_lnkcnt_get(void *node)
472
{
473
	return ((fat_node_t *)node)->lnkcnt;
474
}
475
 
2845 jermar 476
static bool fat_has_children(void *node)
477
{
478
	fat_node_t *nodep = (fat_node_t *)node;
479
	unsigned bps;
480
	unsigned dps;
481
	unsigned blocks;
482
	block_t *b;
483
	unsigned i, j;
484
 
485
	if (nodep->type != FAT_DIRECTORY)
486
		return false;
487
 
2951 jermar 488
	futex_down(&nodep->idx->lock);
2864 jermar 489
	bps = fat_bps_get(nodep->idx->dev_handle);
2845 jermar 490
	dps = bps / sizeof(fat_dentry_t);
491
 
492
	blocks = nodep->size / bps + (nodep->size % bps != 0);
493
 
494
	for (i = 0; i < blocks; i++) {
495
		unsigned dentries;
496
		fat_dentry_t *d;
497
 
2864 jermar 498
		b = fat_block_get(nodep, i);
2845 jermar 499
		dentries = (i == blocks - 1) ?
500
		    nodep->size % sizeof(fat_dentry_t) :
501
		    dps;
502
		for (j = 0; j < dentries; j++) {
503
			d = ((fat_dentry_t *)b->data) + j;
504
			switch (fat_classify_dentry(d)) {
505
			case FAT_DENTRY_SKIP:
506
				continue;
507
			case FAT_DENTRY_LAST:
508
				block_put(b);
2951 jermar 509
				futex_up(&nodep->idx->lock);
2845 jermar 510
				return false;
511
			default:
512
			case FAT_DENTRY_VALID:
513
				block_put(b);
2951 jermar 514
				futex_up(&nodep->idx->lock);
2845 jermar 515
				return true;
516
			}
517
			block_put(b);
2951 jermar 518
			futex_up(&nodep->idx->lock);
2845 jermar 519
			return true;
520
		}
521
		block_put(b);
522
	}
523
 
2951 jermar 524
	futex_up(&nodep->idx->lock);
2845 jermar 525
	return false;
526
}
527
 
2844 jermar 528
static void *fat_root_get(dev_handle_t dev_handle)
529
{
3119 jermar 530
	return fat_node_get(dev_handle, 0);
2844 jermar 531
}
532
 
533
static char fat_plb_get_char(unsigned pos)
534
{
535
	return fat_reg.plb_ro[pos % PLB_SIZE];
536
}
537
 
2831 jermar 538
static bool fat_is_directory(void *node)
539
{
540
	return ((fat_node_t *)node)->type == FAT_DIRECTORY;
541
}
542
 
543
static bool fat_is_file(void *node)
544
{
545
	return ((fat_node_t *)node)->type == FAT_FILE;
546
}
547
 
2793 jermar 548
/** libfs operations */
549
libfs_ops_t fat_libfs_ops = {
550
	.match = fat_match,
551
	.node_get = fat_node_get,
2852 jermar 552
	.node_put = fat_node_put,
2857 jermar 553
	.create = fat_create,
554
	.destroy = fat_destroy,
555
	.link = fat_link,
556
	.unlink = fat_unlink,
2831 jermar 557
	.index_get = fat_index_get,
558
	.size_get = fat_size_get,
559
	.lnkcnt_get = fat_lnkcnt_get,
2845 jermar 560
	.has_children = fat_has_children,
2844 jermar 561
	.root_get = fat_root_get,
562
	.plb_get_char =	fat_plb_get_char,
2831 jermar 563
	.is_directory = fat_is_directory,
564
	.is_file = fat_is_file
2793 jermar 565
};
566
 
3110 jermar 567
void fat_mounted(ipc_callid_t rid, ipc_call_t *request)
568
{
569
	dev_handle_t dev_handle = (dev_handle_t) IPC_GET_ARG1(*request);
3119 jermar 570
	block_t *bb;
571
	uint16_t rde;
3110 jermar 572
	int rc;
573
 
3119 jermar 574
	/* Read the number of root directory entries. */
3253 jermar 575
	bb = block_get(dev_handle, BS_BLOCK, BS_SIZE);
3119 jermar 576
	rde = uint16_t_le2host(FAT_BS(bb)->root_ent_max);
577
	block_put(bb);
578
 
3110 jermar 579
	rc = fat_idx_init_by_dev_handle(dev_handle);
580
	if (rc != EOK) {
581
		ipc_answer_0(rid, rc);
582
		return;
583
	}
584
 
3119 jermar 585
	/* Initialize the root node. */
586
	fat_node_t *rootp = (fat_node_t *)malloc(sizeof(fat_node_t));
587
	if (!rootp) {
588
		fat_idx_fini_by_dev_handle(dev_handle);
589
		ipc_answer_0(rid, ENOMEM);
590
		return;
591
	}
592
	fat_node_initialize(rootp);
593
 
594
	fat_idx_t *ridxp = fat_idx_get_by_pos(dev_handle, FAT_CLST_ROOTPAR, 0);
595
	if (!ridxp) {
596
		free(rootp);
597
		fat_idx_fini_by_dev_handle(dev_handle);
598
		ipc_answer_0(rid, ENOMEM);
599
		return;
600
	}
601
	assert(ridxp->index == 0);
602
	/* ridxp->lock held */
603
 
604
	rootp->type = FAT_DIRECTORY;
605
	rootp->firstc = FAT_CLST_ROOT;
606
	rootp->refcnt = 1;
607
	rootp->size = rde * sizeof(fat_dentry_t);
608
	rootp->idx = ridxp;
609
	ridxp->nodep = rootp;
610
 
611
	futex_up(&ridxp->lock);
612
 
3110 jermar 613
	ipc_answer_0(rid, EOK);
614
}
615
 
616
void fat_mount(ipc_callid_t rid, ipc_call_t *request)
617
{
618
	ipc_answer_0(rid, ENOTSUP);
619
}
620
 
2627 jermar 621
void fat_lookup(ipc_callid_t rid, ipc_call_t *request)
622
{
2793 jermar 623
	libfs_lookup(&fat_libfs_ops, fat_reg.fs_handle, rid, request);
2627 jermar 624
}
625
 
626
/**
627
 * @}
628
 */