pandora: reserve CMA area for c64_tools
[pandora-kernel.git] / fs / jffs2 / build.c
1 /*
2  * JFFS2 -- Journalling Flash File System, Version 2.
3  *
4  * Copyright © 2001-2007 Red Hat, Inc.
5  * Copyright © 2004-2010 David Woodhouse <dwmw2@infradead.org>
6  *
7  * Created by David Woodhouse <dwmw2@infradead.org>
8  *
9  * For licensing information, see the file 'LICENCE' in this directory.
10  *
11  */
12
13 #include <linux/kernel.h>
14 #include <linux/sched.h>
15 #include <linux/slab.h>
16 #include <linux/vmalloc.h>
17 #include <linux/mtd/mtd.h>
18 #include "nodelist.h"
19
20 static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *,
21                 struct jffs2_inode_cache *, struct jffs2_full_dirent **);
22
23 static inline struct jffs2_inode_cache *
24 first_inode_chain(int *i, struct jffs2_sb_info *c)
25 {
26         for (; *i < c->inocache_hashsize; (*i)++) {
27                 if (c->inocache_list[*i])
28                         return c->inocache_list[*i];
29         }
30         return NULL;
31 }
32
33 static inline struct jffs2_inode_cache *
34 next_inode(int *i, struct jffs2_inode_cache *ic, struct jffs2_sb_info *c)
35 {
36         /* More in this chain? */
37         if (ic->next)
38                 return ic->next;
39         (*i)++;
40         return first_inode_chain(i, c);
41 }
42
43 #define for_each_inode(i, c, ic)                        \
44         for (i = 0, ic = first_inode_chain(&i, (c));    \
45              ic;                                        \
46              ic = next_inode(&i, ic, (c)))
47
48
49 static void jffs2_build_inode_pass1(struct jffs2_sb_info *c,
50                                     struct jffs2_inode_cache *ic)
51 {
52         struct jffs2_full_dirent *fd;
53
54         dbg_fsbuild("building directory inode #%u\n", ic->ino);
55
56         /* For each child, increase nlink */
57         for(fd = ic->scan_dents; fd; fd = fd->next) {
58                 struct jffs2_inode_cache *child_ic;
59                 if (!fd->ino)
60                         continue;
61
62                 /* we can get high latency here with huge directories */
63
64                 child_ic = jffs2_get_ino_cache(c, fd->ino);
65                 if (!child_ic) {
66                         dbg_fsbuild("child \"%s\" (ino #%u) of dir ino #%u doesn't exist!\n",
67                                   fd->name, fd->ino, ic->ino);
68                         jffs2_mark_node_obsolete(c, fd->raw);
69                         continue;
70                 }
71
72                 if (fd->type == DT_DIR) {
73                         if (child_ic->pino_nlink) {
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                         } else {
78                                 child_ic->pino_nlink = ic->ino;
79                         }
80                 } else
81                         child_ic->pino_nlink++;
82
83                 dbg_fsbuild("increased nlink for child \"%s\" (ino #%u)\n", fd->name, fd->ino);
84                 /* Can't free scan_dents so far. We might need them in pass 2 */
85         }
86 }
87
88 /* Scan plan:
89  - Scan physical nodes. Build map of inodes/dirents. Allocate inocaches as we go
90  - Scan directory tree from top down, setting nlink in inocaches
91  - Scan inocaches for inodes with nlink==0
92 */
93 static int jffs2_build_filesystem(struct jffs2_sb_info *c)
94 {
95         int ret;
96         int i;
97         struct jffs2_inode_cache *ic;
98         struct jffs2_full_dirent *fd;
99         struct jffs2_full_dirent *dead_fds = NULL;
100
101         dbg_fsbuild("build FS data structures\n");
102
103         /* First, scan the medium and build all the inode caches with
104            lists of physical nodes */
105
106         c->flags |= JFFS2_SB_FLAG_SCANNING;
107         ret = jffs2_scan_medium(c);
108         c->flags &= ~JFFS2_SB_FLAG_SCANNING;
109         if (ret)
110                 goto exit;
111
112         dbg_fsbuild("scanned flash completely\n");
113         jffs2_dbg_dump_block_lists_nolock(c);
114
115         dbg_fsbuild("pass 1 starting\n");
116         c->flags |= JFFS2_SB_FLAG_BUILDING;
117         /* Now scan the directory tree, increasing nlink according to every dirent found. */
118         for_each_inode(i, c, ic) {
119                 if (ic->scan_dents) {
120                         jffs2_build_inode_pass1(c, ic);
121                         cond_resched();
122                 }
123         }
124
125         dbg_fsbuild("pass 1 complete\n");
126
127         /* Next, scan for inodes with nlink == 0 and remove them. If
128            they were directories, then decrement the nlink of their
129            children too, and repeat the scan. As that's going to be
130            a fairly uncommon occurrence, it's not so evil to do it this
131            way. Recursion bad. */
132         dbg_fsbuild("pass 2 starting\n");
133
134         for_each_inode(i, c, ic) {
135                 if (ic->pino_nlink)
136                         continue;
137
138                 jffs2_build_remove_unlinked_inode(c, ic, &dead_fds);
139                 cond_resched();
140         }
141
142         dbg_fsbuild("pass 2a starting\n");
143
144         while (dead_fds) {
145                 fd = dead_fds;
146                 dead_fds = fd->next;
147
148                 ic = jffs2_get_ino_cache(c, fd->ino);
149
150                 if (ic)
151                         jffs2_build_remove_unlinked_inode(c, ic, &dead_fds);
152                 jffs2_free_full_dirent(fd);
153         }
154
155         dbg_fsbuild("pass 2a complete\n");
156         dbg_fsbuild("freeing temporary data structures\n");
157
158         /* Finally, we can scan again and free the dirent structs */
159         for_each_inode(i, c, ic) {
160                 while(ic->scan_dents) {
161                         fd = ic->scan_dents;
162                         ic->scan_dents = fd->next;
163                         jffs2_free_full_dirent(fd);
164                 }
165                 ic->scan_dents = NULL;
166                 cond_resched();
167         }
168         jffs2_build_xattr_subsystem(c);
169         c->flags &= ~JFFS2_SB_FLAG_BUILDING;
170
171         dbg_fsbuild("FS build complete\n");
172
173         /* Rotate the lists by some number to ensure wear levelling */
174         jffs2_rotate_lists(c);
175
176         ret = 0;
177
178 exit:
179         if (ret) {
180                 for_each_inode(i, c, ic) {
181                         while(ic->scan_dents) {
182                                 fd = ic->scan_dents;
183                                 ic->scan_dents = fd->next;
184                                 jffs2_free_full_dirent(fd);
185                         }
186                 }
187                 jffs2_clear_xattr_subsystem(c);
188         }
189
190         return ret;
191 }
192
193 static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *c,
194                                         struct jffs2_inode_cache *ic,
195                                         struct jffs2_full_dirent **dead_fds)
196 {
197         struct jffs2_raw_node_ref *raw;
198         struct jffs2_full_dirent *fd;
199
200         dbg_fsbuild("removing ino #%u with nlink == zero.\n", ic->ino);
201
202         raw = ic->nodes;
203         while (raw != (void *)ic) {
204                 struct jffs2_raw_node_ref *next = raw->next_in_ino;
205                 dbg_fsbuild("obsoleting node at 0x%08x\n", ref_offset(raw));
206                 jffs2_mark_node_obsolete(c, raw);
207                 raw = next;
208         }
209
210         if (ic->scan_dents) {
211                 int whinged = 0;
212                 dbg_fsbuild("inode #%u was a directory which may have children...\n", ic->ino);
213
214                 while(ic->scan_dents) {
215                         struct jffs2_inode_cache *child_ic;
216
217                         fd = ic->scan_dents;
218                         ic->scan_dents = fd->next;
219
220                         if (!fd->ino) {
221                                 /* It's a deletion dirent. Ignore it */
222                                 dbg_fsbuild("child \"%s\" is a deletion dirent, skipping...\n", fd->name);
223                                 jffs2_free_full_dirent(fd);
224                                 continue;
225                         }
226                         if (!whinged)
227                                 whinged = 1;
228
229                         dbg_fsbuild("removing child \"%s\", ino #%u\n", fd->name, fd->ino);
230
231                         child_ic = jffs2_get_ino_cache(c, fd->ino);
232                         if (!child_ic) {
233                                 dbg_fsbuild("cannot remove child \"%s\", ino #%u, because it doesn't exist\n",
234                                                 fd->name, fd->ino);
235                                 jffs2_free_full_dirent(fd);
236                                 continue;
237                         }
238
239                         /* Reduce nlink of the child. If it's now zero, stick it on the
240                            dead_fds list to be cleaned up later. Else just free the fd */
241
242                         if (fd->type == DT_DIR)
243                                 child_ic->pino_nlink = 0;
244                         else
245                                 child_ic->pino_nlink--;
246
247                         if (!child_ic->pino_nlink) {
248                                 dbg_fsbuild("inode #%u (\"%s\") now has no links; adding to dead_fds list.\n",
249                                           fd->ino, fd->name);
250                                 fd->next = *dead_fds;
251                                 *dead_fds = fd;
252                         } else {
253                                 dbg_fsbuild("inode #%u (\"%s\") has now got nlink %d. Ignoring.\n",
254                                           fd->ino, fd->name, child_ic->pino_nlink);
255                                 jffs2_free_full_dirent(fd);
256                         }
257                 }
258         }
259
260         /*
261            We don't delete the inocache from the hash list and free it yet.
262            The erase code will do that, when all the nodes are completely gone.
263         */
264 }
265
266 static void jffs2_calc_trigger_levels(struct jffs2_sb_info *c)
267 {
268         uint32_t size;
269
270         /* Deletion should almost _always_ be allowed. We're fairly
271            buggered once we stop allowing people to delete stuff
272            because there's not enough free space... */
273         c->resv_blocks_deletion = 2;
274
275         /* Be conservative about how much space we need before we allow writes.
276            On top of that which is required for deletia, require an extra 2%
277            of the medium to be available, for overhead caused by nodes being
278            split across blocks, etc. */
279
280         size = c->flash_size / 50; /* 2% of flash size */
281         size += c->nr_blocks * 100; /* And 100 bytes per eraseblock */
282         size += c->sector_size - 1; /* ... and round up */
283
284         c->resv_blocks_write = c->resv_blocks_deletion + (size / c->sector_size);
285
286         /* When do we let the GC thread run in the background */
287
288         c->resv_blocks_gctrigger = c->resv_blocks_write + 1;
289
290         /* When do we allow garbage collection to merge nodes to make
291            long-term progress at the expense of short-term space exhaustion? */
292         c->resv_blocks_gcmerge = c->resv_blocks_deletion + 1;
293
294         /* When do we allow garbage collection to eat from bad blocks rather
295            than actually making progress? */
296         c->resv_blocks_gcbad = 0;//c->resv_blocks_deletion + 2;
297
298         /* What number of 'very dirty' eraseblocks do we allow before we
299            trigger the GC thread even if we don't _need_ the space. When we
300            can't mark nodes obsolete on the medium, the old dirty nodes cause
301            performance problems because we have to inspect and discard them. */
302         c->vdirty_blocks_gctrigger = c->resv_blocks_gctrigger;
303         if (jffs2_can_mark_obsolete(c))
304                 c->vdirty_blocks_gctrigger *= 10;
305
306         /* If there's less than this amount of dirty space, don't bother
307            trying to GC to make more space. It'll be a fruitless task */
308         c->nospc_dirty_size = c->sector_size + (c->flash_size / 100);
309
310         dbg_fsbuild("JFFS2 trigger levels (size %d KiB, block size %d KiB, %d blocks)\n",
311                   c->flash_size / 1024, c->sector_size / 1024, c->nr_blocks);
312         dbg_fsbuild("Blocks required to allow deletion:    %d (%d KiB)\n",
313                   c->resv_blocks_deletion, c->resv_blocks_deletion*c->sector_size/1024);
314         dbg_fsbuild("Blocks required to allow writes:      %d (%d KiB)\n",
315                   c->resv_blocks_write, c->resv_blocks_write*c->sector_size/1024);
316         dbg_fsbuild("Blocks required to quiesce GC thread: %d (%d KiB)\n",
317                   c->resv_blocks_gctrigger, c->resv_blocks_gctrigger*c->sector_size/1024);
318         dbg_fsbuild("Blocks required to allow GC merges:   %d (%d KiB)\n",
319                   c->resv_blocks_gcmerge, c->resv_blocks_gcmerge*c->sector_size/1024);
320         dbg_fsbuild("Blocks required to GC bad blocks:     %d (%d KiB)\n",
321                   c->resv_blocks_gcbad, c->resv_blocks_gcbad*c->sector_size/1024);
322         dbg_fsbuild("Amount of dirty space required to GC: %d bytes\n",
323                   c->nospc_dirty_size);
324         dbg_fsbuild("Very dirty blocks before GC triggered: %d\n",
325                   c->vdirty_blocks_gctrigger);
326 }
327
328 int jffs2_do_mount_fs(struct jffs2_sb_info *c)
329 {
330         int ret;
331         int i;
332         int size;
333
334         c->free_size = c->flash_size;
335         c->nr_blocks = c->flash_size / c->sector_size;
336         size = sizeof(struct jffs2_eraseblock) * c->nr_blocks;
337 #ifndef __ECOS
338         if (jffs2_blocks_use_vmalloc(c))
339                 c->blocks = vzalloc(size);
340         else
341 #endif
342                 c->blocks = kzalloc(size, GFP_KERNEL);
343         if (!c->blocks)
344                 return -ENOMEM;
345
346         for (i=0; i<c->nr_blocks; i++) {
347                 INIT_LIST_HEAD(&c->blocks[i].list);
348                 c->blocks[i].offset = i * c->sector_size;
349                 c->blocks[i].free_size = c->sector_size;
350         }
351
352         INIT_LIST_HEAD(&c->clean_list);
353         INIT_LIST_HEAD(&c->very_dirty_list);
354         INIT_LIST_HEAD(&c->dirty_list);
355         INIT_LIST_HEAD(&c->erasable_list);
356         INIT_LIST_HEAD(&c->erasing_list);
357         INIT_LIST_HEAD(&c->erase_checking_list);
358         INIT_LIST_HEAD(&c->erase_pending_list);
359         INIT_LIST_HEAD(&c->erasable_pending_wbuf_list);
360         INIT_LIST_HEAD(&c->erase_complete_list);
361         INIT_LIST_HEAD(&c->free_list);
362         INIT_LIST_HEAD(&c->bad_list);
363         INIT_LIST_HEAD(&c->bad_used_list);
364         c->highest_ino = 1;
365         c->summary = NULL;
366
367         ret = jffs2_sum_init(c);
368         if (ret)
369                 goto out_free;
370
371         if (jffs2_build_filesystem(c)) {
372                 dbg_fsbuild("build_fs failed\n");
373                 jffs2_free_ino_caches(c);
374                 jffs2_free_raw_node_refs(c);
375                 ret = -EIO;
376                 goto out_free;
377         }
378
379         jffs2_calc_trigger_levels(c);
380
381         return 0;
382
383  out_free:
384 #ifndef __ECOS
385         if (jffs2_blocks_use_vmalloc(c))
386                 vfree(c->blocks);
387         else
388 #endif
389                 kfree(c->blocks);
390
391         return ret;
392 }