Merge git://git.infradead.org/mtd-2.6
[linux-3.10.git] / fs / jffs2 / build.c
1 /*
2  * JFFS2 -- Journalling Flash File System, Version 2.
3  *
4  * Copyright (C) 2001-2003 Red Hat, Inc.
5  *
6  * Created by David Woodhouse <dwmw2@infradead.org>
7  *
8  * For licensing information, see the file 'LICENCE' in this directory.
9  *
10  * $Id: build.c,v 1.85 2005/11/07 11:14:38 gleixner Exp $
11  *
12  */
13
14 #include <linux/kernel.h>
15 #include <linux/sched.h>
16 #include <linux/slab.h>
17 #include <linux/vmalloc.h>
18 #include <linux/mtd/mtd.h>
19 #include "nodelist.h"
20
21 static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *,
22                 struct jffs2_inode_cache *, struct jffs2_full_dirent **);
23
24 static inline struct jffs2_inode_cache *
25 first_inode_chain(int *i, struct jffs2_sb_info *c)
26 {
27         for (; *i < INOCACHE_HASHSIZE; (*i)++) {
28                 if (c->inocache_list[*i])
29                         return c->inocache_list[*i];
30         }
31         return NULL;
32 }
33
34 static inline struct jffs2_inode_cache *
35 next_inode(int *i, struct jffs2_inode_cache *ic, struct jffs2_sb_info *c)
36 {
37         /* More in this chain? */
38         if (ic->next)
39                 return ic->next;
40         (*i)++;
41         return first_inode_chain(i, c);
42 }
43
44 #define for_each_inode(i, c, ic)                        \
45         for (i = 0, ic = first_inode_chain(&i, (c));    \
46              ic;                                        \
47              ic = next_inode(&i, ic, (c)))
48
49
50 static void jffs2_build_inode_pass1(struct jffs2_sb_info *c,
51                                         struct jffs2_inode_cache *ic)
52 {
53         struct jffs2_full_dirent *fd;
54
55         dbg_fsbuild("building directory inode #%u\n", ic->ino);
56
57         /* For each child, increase nlink */
58         for(fd = ic->scan_dents; fd; fd = fd->next) {
59                 struct jffs2_inode_cache *child_ic;
60                 if (!fd->ino)
61                         continue;
62
63                 /* we can get high latency here with huge directories */
64
65                 child_ic = jffs2_get_ino_cache(c, fd->ino);
66                 if (!child_ic) {
67                         dbg_fsbuild("child \"%s\" (ino #%u) of dir ino #%u doesn't exist!\n",
68                                   fd->name, fd->ino, ic->ino);
69                         jffs2_mark_node_obsolete(c, fd->raw);
70                         continue;
71                 }
72
73                 if (child_ic->nlink++ && fd->type == DT_DIR) {
74                         JFFS2_ERROR("child dir \"%s\" (ino #%u) of dir ino #%u appears to be a hard link\n",
75                                 fd->name, fd->ino, ic->ino);
76                         /* TODO: What do we do about it? */
77                 }
78                 dbg_fsbuild("increased nlink for child \"%s\" (ino #%u)\n", fd->name, fd->ino);
79                 /* Can't free scan_dents so far. We might need them in pass 2 */
80         }
81 }
82
83 /* Scan plan:
84  - Scan physical nodes. Build map of inodes/dirents. Allocate inocaches as we go
85  - Scan directory tree from top down, setting nlink in inocaches
86  - Scan inocaches for inodes with nlink==0
87 */
88 static int jffs2_build_filesystem(struct jffs2_sb_info *c)
89 {
90         int ret;
91         int i;
92         struct jffs2_inode_cache *ic;
93         struct jffs2_full_dirent *fd;
94         struct jffs2_full_dirent *dead_fds = NULL;
95
96         dbg_fsbuild("build FS data structures\n");
97
98         /* First, scan the medium and build all the inode caches with
99            lists of physical nodes */
100
101         c->flags |= JFFS2_SB_FLAG_SCANNING;
102         ret = jffs2_scan_medium(c);
103         c->flags &= ~JFFS2_SB_FLAG_SCANNING;
104         if (ret)
105                 goto exit;
106
107         dbg_fsbuild("scanned flash completely\n");
108         jffs2_dbg_dump_block_lists_nolock(c);
109
110         dbg_fsbuild("pass 1 starting\n");
111         c->flags |= JFFS2_SB_FLAG_BUILDING;
112         /* Now scan the directory tree, increasing nlink according to every dirent found. */
113         for_each_inode(i, c, ic) {
114                 if (ic->scan_dents) {
115                         jffs2_build_inode_pass1(c, ic);
116                         cond_resched();
117                 }
118         }
119
120         dbg_fsbuild("pass 1 complete\n");
121
122         /* Next, scan for inodes with nlink == 0 and remove them. If
123            they were directories, then decrement the nlink of their
124            children too, and repeat the scan. As that's going to be
125            a fairly uncommon occurrence, it's not so evil to do it this
126            way. Recursion bad. */
127         dbg_fsbuild("pass 2 starting\n");
128
129         for_each_inode(i, c, ic) {
130                 if (ic->nlink)
131                         continue;
132
133                 jffs2_build_remove_unlinked_inode(c, ic, &dead_fds);
134                 cond_resched();
135         }
136
137         dbg_fsbuild("pass 2a starting\n");
138
139         while (dead_fds) {
140                 fd = dead_fds;
141                 dead_fds = fd->next;
142
143                 ic = jffs2_get_ino_cache(c, fd->ino);
144
145                 if (ic)
146                         jffs2_build_remove_unlinked_inode(c, ic, &dead_fds);
147                 jffs2_free_full_dirent(fd);
148         }
149
150         dbg_fsbuild("pass 2a complete\n");
151         dbg_fsbuild("freeing temporary data structures\n");
152
153         /* Finally, we can scan again and free the dirent structs */
154         for_each_inode(i, c, ic) {
155                 while(ic->scan_dents) {
156                         fd = ic->scan_dents;
157                         ic->scan_dents = fd->next;
158                         jffs2_free_full_dirent(fd);
159                 }
160                 ic->scan_dents = NULL;
161                 cond_resched();
162         }
163         jffs2_build_xattr_subsystem(c);
164         c->flags &= ~JFFS2_SB_FLAG_BUILDING;
165
166         dbg_fsbuild("FS build complete\n");
167
168         /* Rotate the lists by some number to ensure wear levelling */
169         jffs2_rotate_lists(c);
170
171         ret = 0;
172
173 exit:
174         if (ret) {
175                 for_each_inode(i, c, ic) {
176                         while(ic->scan_dents) {
177                                 fd = ic->scan_dents;
178                                 ic->scan_dents = fd->next;
179                                 jffs2_free_full_dirent(fd);
180                         }
181                 }
182                 jffs2_clear_xattr_subsystem(c);
183         }
184
185         return ret;
186 }
187
188 static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *c,
189                                         struct jffs2_inode_cache *ic,
190                                         struct jffs2_full_dirent **dead_fds)
191 {
192         struct jffs2_raw_node_ref *raw;
193         struct jffs2_full_dirent *fd;
194
195         dbg_fsbuild("removing ino #%u with nlink == zero.\n", ic->ino);
196
197         raw = ic->nodes;
198         while (raw != (void *)ic) {
199                 struct jffs2_raw_node_ref *next = raw->next_in_ino;
200                 dbg_fsbuild("obsoleting node at 0x%08x\n", ref_offset(raw));
201                 jffs2_mark_node_obsolete(c, raw);
202                 raw = next;
203         }
204
205         if (ic->scan_dents) {
206                 int whinged = 0;
207                 dbg_fsbuild("inode #%u was a directory which may have children...\n", ic->ino);
208
209                 while(ic->scan_dents) {
210                         struct jffs2_inode_cache *child_ic;
211
212                         fd = ic->scan_dents;
213                         ic->scan_dents = fd->next;
214
215                         if (!fd->ino) {
216                                 /* It's a deletion dirent. Ignore it */
217                                 dbg_fsbuild("child \"%s\" is a deletion dirent, skipping...\n", fd->name);
218                                 jffs2_free_full_dirent(fd);
219                                 continue;
220                         }
221                         if (!whinged)
222                                 whinged = 1;
223
224                         dbg_fsbuild("removing child \"%s\", ino #%u\n", fd->name, fd->ino);
225
226                         child_ic = jffs2_get_ino_cache(c, fd->ino);
227                         if (!child_ic) {
228                                 dbg_fsbuild("cannot remove child \"%s\", ino #%u, because it doesn't exist\n",
229                                                 fd->name, fd->ino);
230                                 jffs2_free_full_dirent(fd);
231                                 continue;
232                         }
233
234                         /* Reduce nlink of the child. If it's now zero, stick it on the
235                            dead_fds list to be cleaned up later. Else just free the fd */
236
237                         child_ic->nlink--;
238
239                         if (!child_ic->nlink) {
240                                 dbg_fsbuild("inode #%u (\"%s\") has now got zero nlink, adding to dead_fds list.\n",
241                                           fd->ino, fd->name);
242                                 fd->next = *dead_fds;
243                                 *dead_fds = fd;
244                         } else {
245                                 dbg_fsbuild("inode #%u (\"%s\") has now got nlink %d. Ignoring.\n",
246                                           fd->ino, fd->name, child_ic->nlink);
247                                 jffs2_free_full_dirent(fd);
248                         }
249                 }
250         }
251
252         /*
253            We don't delete the inocache from the hash list and free it yet.
254            The erase code will do that, when all the nodes are completely gone.
255         */
256 }
257
258 static void jffs2_calc_trigger_levels(struct jffs2_sb_info *c)
259 {
260         uint32_t size;
261
262         /* Deletion should almost _always_ be allowed. We're fairly
263            buggered once we stop allowing people to delete stuff
264            because there's not enough free space... */
265         c->resv_blocks_deletion = 2;
266
267         /* Be conservative about how much space we need before we allow writes.
268            On top of that which is required for deletia, require an extra 2%
269            of the medium to be available, for overhead caused by nodes being
270            split across blocks, etc. */
271
272         size = c->flash_size / 50; /* 2% of flash size */
273         size += c->nr_blocks * 100; /* And 100 bytes per eraseblock */
274         size += c->sector_size - 1; /* ... and round up */
275
276         c->resv_blocks_write = c->resv_blocks_deletion + (size / c->sector_size);
277
278         /* When do we let the GC thread run in the background */
279
280         c->resv_blocks_gctrigger = c->resv_blocks_write + 1;
281
282         /* When do we allow garbage collection to merge nodes to make
283            long-term progress at the expense of short-term space exhaustion? */
284         c->resv_blocks_gcmerge = c->resv_blocks_deletion + 1;
285
286         /* When do we allow garbage collection to eat from bad blocks rather
287            than actually making progress? */
288         c->resv_blocks_gcbad = 0;//c->resv_blocks_deletion + 2;
289
290         /* If there's less than this amount of dirty space, don't bother
291            trying to GC to make more space. It'll be a fruitless task */
292         c->nospc_dirty_size = c->sector_size + (c->flash_size / 100);
293
294         dbg_fsbuild("JFFS2 trigger levels (size %d KiB, block size %d KiB, %d blocks)\n",
295                   c->flash_size / 1024, c->sector_size / 1024, c->nr_blocks);
296         dbg_fsbuild("Blocks required to allow deletion:    %d (%d KiB)\n",
297                   c->resv_blocks_deletion, c->resv_blocks_deletion*c->sector_size/1024);
298         dbg_fsbuild("Blocks required to allow writes:      %d (%d KiB)\n",
299                   c->resv_blocks_write, c->resv_blocks_write*c->sector_size/1024);
300         dbg_fsbuild("Blocks required to quiesce GC thread: %d (%d KiB)\n",
301                   c->resv_blocks_gctrigger, c->resv_blocks_gctrigger*c->sector_size/1024);
302         dbg_fsbuild("Blocks required to allow GC merges:   %d (%d KiB)\n",
303                   c->resv_blocks_gcmerge, c->resv_blocks_gcmerge*c->sector_size/1024);
304         dbg_fsbuild("Blocks required to GC bad blocks:     %d (%d KiB)\n",
305                   c->resv_blocks_gcbad, c->resv_blocks_gcbad*c->sector_size/1024);
306         dbg_fsbuild("Amount of dirty space required to GC: %d bytes\n",
307                   c->nospc_dirty_size);
308 }
309
310 int jffs2_do_mount_fs(struct jffs2_sb_info *c)
311 {
312         int ret;
313         int i;
314         int size;
315
316         c->free_size = c->flash_size;
317         c->nr_blocks = c->flash_size / c->sector_size;
318         size = sizeof(struct jffs2_eraseblock) * c->nr_blocks;
319 #ifndef __ECOS
320         if (jffs2_blocks_use_vmalloc(c))
321                 c->blocks = vmalloc(size);
322         else
323 #endif
324                 c->blocks = kmalloc(size, GFP_KERNEL);
325         if (!c->blocks)
326                 return -ENOMEM;
327
328         memset(c->blocks, 0, size);
329         for (i=0; i<c->nr_blocks; i++) {
330                 INIT_LIST_HEAD(&c->blocks[i].list);
331                 c->blocks[i].offset = i * c->sector_size;
332                 c->blocks[i].free_size = c->sector_size;
333         }
334
335         INIT_LIST_HEAD(&c->clean_list);
336         INIT_LIST_HEAD(&c->very_dirty_list);
337         INIT_LIST_HEAD(&c->dirty_list);
338         INIT_LIST_HEAD(&c->erasable_list);
339         INIT_LIST_HEAD(&c->erasing_list);
340         INIT_LIST_HEAD(&c->erase_pending_list);
341         INIT_LIST_HEAD(&c->erasable_pending_wbuf_list);
342         INIT_LIST_HEAD(&c->erase_complete_list);
343         INIT_LIST_HEAD(&c->free_list);
344         INIT_LIST_HEAD(&c->bad_list);
345         INIT_LIST_HEAD(&c->bad_used_list);
346         c->highest_ino = 1;
347         c->summary = NULL;
348
349         ret = jffs2_sum_init(c);
350         if (ret)
351                 return ret;
352
353         if (jffs2_build_filesystem(c)) {
354                 dbg_fsbuild("build_fs failed\n");
355                 jffs2_free_ino_caches(c);
356                 jffs2_free_raw_node_refs(c);
357 #ifndef __ECOS
358                 if (jffs2_blocks_use_vmalloc(c))
359                         vfree(c->blocks);
360                 else
361 #endif
362                         kfree(c->blocks);
363
364                 return -EIO;
365         }
366
367         jffs2_calc_trigger_levels(c);
368
369         return 0;
370 }