1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
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
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
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
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
|
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright 2017 Omnibond Systems, L.L.C.
*/
#include "protocol.h"
#include "orangefs-kernel.h"
#include "orangefs-bufmap.h"
struct orangefs_dir_part {
struct orangefs_dir_part *next;
size_t len;
};
struct orangefs_dir {
__u64 token;
struct orangefs_dir_part *part;
loff_t end;
int error;
};
#define PART_SHIFT (24)
#define PART_SIZE (1<<24)
#define PART_MASK (~(PART_SIZE - 1))
/*
* There can be up to 512 directory entries. Each entry is encoded as
* follows:
* 4 bytes: string size (n)
* n bytes: string
* 1 byte: trailing zero
* padding to 8 bytes
* 16 bytes: khandle
* padding to 8 bytes
*
* The trailer_buf starts with a struct orangefs_readdir_response_s
* which must be skipped to get to the directory data.
*
* The data which is received from the userspace daemon is termed a
* part and is stored in a linked list in case more than one part is
* needed for a large directory.
*
* The position pointer (ctx->pos) encodes the part and offset on which
* to begin reading at. Bits above PART_SHIFT encode the part and bits
* below PART_SHIFT encode the offset. Parts are stored in a linked
* list which grows as data is received from the server. The overhead
* associated with managing the list is presumed to be small compared to
* the overhead of communicating with the server.
*
* As data is received from the server, it is placed at the end of the
* part list. Data is parsed from the current position as it is needed.
* When data is determined to be corrupt, it is either because the
* userspace component has sent back corrupt data or because the file
* pointer has been moved to an invalid location. Since the two cannot
* be differentiated, return EIO.
*
* Part zero is synthesized to contains `.' and `..'. Part one is the
* first part of the part list.
*/
static int do_readdir(struct orangefs_inode_s *oi,
struct orangefs_dir *od, struct dentry *dentry,
struct orangefs_kernel_op_s *op)
{
struct orangefs_readdir_response_s *resp;
int bufi, r;
/*
* Despite the badly named field, readdir does not use shared
* memory. However, there are a limited number of readdir
* slots, which must be allocated here. This flag simply tells
* the op scheduler to return the op here for retry.
*/
op->uses_shared_memory = 1;
op->upcall.req.readdir.refn = oi->refn;
op->upcall.req.readdir.token = od->token;
op->upcall.req.readdir.max_dirent_count =
ORANGEFS_MAX_DIRENT_COUNT_READDIR;
again:
bufi = orangefs_readdir_index_get();
if (bufi < 0) {
od->error = bufi;
return bufi;
}
op->upcall.req.readdir.buf_index = bufi;
r = service_operation(op, "orangefs_readdir",
get_interruptible_flag(dentry->d_inode));
orangefs_readdir_index_put(bufi);
if (op_state_purged(op)) {
if (r == -EAGAIN) {
vfree(op->downcall.trailer_buf);
goto again;
} else if (r == -EIO) {
vfree(op->downcall.trailer_buf);
od->error = r;
return r;
}
}
if (r < 0) {
vfree(op->downcall.trailer_buf);
od->error = r;
return r;
} else if (op->downcall.status) {
vfree(op->downcall.trailer_buf);
od->error = op->downcall.status;
return op->downcall.status;
}
/*
* The maximum size is size per entry times the 512 entries plus
* the header. This is well under the limit.
*/
if (op->downcall.trailer_size > PART_SIZE) {
vfree(op->downcall.trailer_buf);
od->error = -EIO;
return -EIO;
}
resp = (struct orangefs_readdir_response_s *)
op->downcall.trailer_buf;
od->token = resp->token;
return 0;
}
static int parse_readdir(struct orangefs_dir *od,
struct orangefs_kernel_op_s *op)
{
struct orangefs_dir_part *part, *new;
size_t count;
count = 1;
part = od->part;
while (part) {
count++;
if (part->next)
part = part->next;
else
break;
}
new = (void *)op->downcall.trailer_buf;
new->next = NULL;
new->len = op->downcall.trailer_size -
sizeof(struct orangefs_readdir_response_s);
if (!od->part)
od->part = new;
else
part->next = new;
count++;
od->end = count << PART_SHIFT;
return 0;
}
static int orangefs_dir_more(struct orangefs_inode_s *oi,
struct orangefs_dir *od, struct dentry *dentry)
{
struct orangefs_kernel_op_s *op;
int r;
op = op_alloc(ORANGEFS_VFS_OP_READDIR);
if (!op) {
od->error = -ENOMEM;
return -ENOMEM;
}
r = do_readdir(oi, od, dentry, op);
if (r) {
od->error = r;
goto out;
}
r = parse_readdir(od, op);
if (r) {
od->error = r;
goto out;
}
od->error = 0;
out:
op_release(op);
return od->error;
}
static int fill_from_part(struct orangefs_dir_part *part,
struct dir_context *ctx)
{
const int offset = sizeof(struct orangefs_readdir_response_s);
struct orangefs_khandle *khandle;
__u32 *len, padlen;
loff_t i;
char *s;
i = ctx->pos & ~PART_MASK;
/* The file offset from userspace is too large. */
if (i > part->len)
return 1;
/*
* If the seek pointer is positioned just before an entry it
* should find the next entry.
*/
if (i % 8)
i = i + (8 - i%8)%8;
while (i < part->len) {
if (part->len < i + sizeof *len)
break;
len = (void *)part + offset + i;
/*
* len is the size of the string itself. padlen is the
* total size of the encoded string.
*/
padlen = (sizeof *len + *len + 1) +
(8 - (sizeof *len + *len + 1)%8)%8;
if (part->len < i + padlen + sizeof *khandle)
goto next;
s = (void *)part + offset + i + sizeof *len;
if (s[*len] != 0)
goto next;
khandle = (void *)part + offset + i + padlen;
if (!dir_emit(ctx, s, *len,
orangefs_khandle_to_ino(khandle),
DT_UNKNOWN))
return 0;
i += padlen + sizeof *khandle;
i = i + (8 - i%8)%8;
BUG_ON(i > part->len);
ctx->pos = (ctx->pos & PART_MASK) | i;
continue;
next:
i += 8;
}
return 1;
}
static int orangefs_dir_fill(struct orangefs_inode_s *oi,
struct orangefs_dir *od, struct dentry *dentry,
struct dir_context *ctx)
{
struct orangefs_dir_part *part;
size_t count;
count = ((ctx->pos & PART_MASK) >> PART_SHIFT) - 1;
part = od->part;
while (part->next && count) {
count--;
part = part->next;
}
/* This means the userspace file offset is invalid. */
if (count) {
od->error = -EIO;
return -EIO;
}
while (part && part->len) {
int r;
r = fill_from_part(part, ctx);
if (r < 0) {
od->error = r;
return r;
} else if (r
|