]>
Commit | Line | Data |
---|---|---|
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 | #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt | |
14 | ||
15 | #include <linux/kernel.h> | |
16 | #include <linux/sched.h> | |
17 | #include <linux/slab.h> | |
18 | #include <linux/vmalloc.h> | |
19 | #include <linux/mtd/mtd.h> | |
20 | #include <linux/mm.h> /* kvfree() */ | |
21 | #include "nodelist.h" | |
22 | ||
23 | static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *, | |
24 | struct jffs2_inode_cache *, struct jffs2_full_dirent **); | |
25 | ||
26 | static inline struct jffs2_inode_cache * | |
27 | first_inode_chain(int *i, struct jffs2_sb_info *c) | |
28 | { | |
29 | for (; *i < c->inocache_hashsize; (*i)++) { | |
30 | if (c->inocache_list[*i]) | |
31 | return c->inocache_list[*i]; | |
32 | } | |
33 | return NULL; | |
34 | } | |
35 | ||
36 | static inline struct jffs2_inode_cache * | |
37 | next_inode(int *i, struct jffs2_inode_cache *ic, struct jffs2_sb_info *c) | |
38 | { | |
39 | /* More in this chain? */ | |
40 | if (ic->next) | |
41 | return ic->next; | |
42 | (*i)++; | |
43 | return first_inode_chain(i, c); | |
44 | } | |
45 | ||
46 | #define for_each_inode(i, c, ic) \ | |
47 | for (i = 0, ic = first_inode_chain(&i, (c)); \ | |
48 | ic; \ | |
49 | ic = next_inode(&i, ic, (c))) | |
50 | ||
51 | ||
52 | static void jffs2_build_inode_pass1(struct jffs2_sb_info *c, | |
53 | struct jffs2_inode_cache *ic) | |
54 | { | |
55 | struct jffs2_full_dirent *fd; | |
56 | ||
57 | dbg_fsbuild("building directory inode #%u\n", ic->ino); | |
58 | ||
59 | /* For each child, increase nlink */ | |
60 | for(fd = ic->scan_dents; fd; fd = fd->next) { | |
61 | struct jffs2_inode_cache *child_ic; | |
62 | if (!fd->ino) | |
63 | continue; | |
64 | ||
65 | /* we can get high latency here with huge directories */ | |
66 | ||
67 | child_ic = jffs2_get_ino_cache(c, fd->ino); | |
68 | if (!child_ic) { | |
69 | dbg_fsbuild("child \"%s\" (ino #%u) of dir ino #%u doesn't exist!\n", | |
70 | fd->name, fd->ino, ic->ino); | |
71 | jffs2_mark_node_obsolete(c, fd->raw); | |
72 | continue; | |
73 | } | |
74 | ||
75 | if (fd->type == DT_DIR) { | |
76 | if (child_ic->pino_nlink) { | |
77 | JFFS2_ERROR("child dir \"%s\" (ino #%u) of dir ino #%u appears to be a hard link\n", | |
78 | fd->name, fd->ino, ic->ino); | |
79 | /* TODO: What do we do about it? */ | |
80 | } else { | |
81 | child_ic->pino_nlink = ic->ino; | |
82 | } | |
83 | } else | |
84 | child_ic->pino_nlink++; | |
85 | ||
86 | dbg_fsbuild("increased nlink for child \"%s\" (ino #%u)\n", fd->name, fd->ino); | |
87 | /* Can't free scan_dents so far. We might need them in pass 2 */ | |
88 | } | |
89 | } | |
90 | ||
91 | /* Scan plan: | |
92 | - Scan physical nodes. Build map of inodes/dirents. Allocate inocaches as we go | |
93 | - Scan directory tree from top down, setting nlink in inocaches | |
94 | - Scan inocaches for inodes with nlink==0 | |
95 | */ | |
96 | static int jffs2_build_filesystem(struct jffs2_sb_info *c) | |
97 | { | |
98 | int ret; | |
99 | int i; | |
100 | struct jffs2_inode_cache *ic; | |
101 | struct jffs2_full_dirent *fd; | |
102 | struct jffs2_full_dirent *dead_fds = NULL; | |
103 | ||
104 | dbg_fsbuild("build FS data structures\n"); | |
105 | ||
106 | /* First, scan the medium and build all the inode caches with | |
107 | lists of physical nodes */ | |
108 | ||
109 | c->flags |= JFFS2_SB_FLAG_SCANNING; | |
110 | ret = jffs2_scan_medium(c); | |
111 | c->flags &= ~JFFS2_SB_FLAG_SCANNING; | |
112 | if (ret) | |
113 | goto exit; | |
114 | ||
115 | dbg_fsbuild("scanned flash completely\n"); | |
116 | jffs2_dbg_dump_block_lists_nolock(c); | |
117 | ||
118 | dbg_fsbuild("pass 1 starting\n"); | |
119 | c->flags |= JFFS2_SB_FLAG_BUILDING; | |
120 | /* Now scan the directory tree, increasing nlink according to every dirent found. */ | |
121 | for_each_inode(i, c, ic) { | |
122 | if (ic->scan_dents) { | |
123 | jffs2_build_inode_pass1(c, ic); | |
124 | cond_resched(); | |
125 | } | |
126 | } | |
127 | ||
128 | dbg_fsbuild("pass 1 complete\n"); | |
129 | ||
130 | /* Next, scan for inodes with nlink == 0 and remove them. If | |
131 | they were directories, then decrement the nlink of their | |
132 | children too, and repeat the scan. As that's going to be | |
133 | a fairly uncommon occurrence, it's not so evil to do it this | |
134 | way. Recursion bad. */ | |
135 | dbg_fsbuild("pass 2 starting\n"); | |
136 | ||
137 | for_each_inode(i, c, ic) { | |
138 | if (ic->pino_nlink) | |
139 | continue; | |
140 | ||
141 | jffs2_build_remove_unlinked_inode(c, ic, &dead_fds); | |
142 | cond_resched(); | |
143 | } | |
144 | ||
145 | dbg_fsbuild("pass 2a starting\n"); | |
146 | ||
147 | while (dead_fds) { | |
148 | fd = dead_fds; | |
149 | dead_fds = fd->next; | |
150 | ||
151 | ic = jffs2_get_ino_cache(c, fd->ino); | |
152 | ||
153 | if (ic) | |
154 | jffs2_build_remove_unlinked_inode(c, ic, &dead_fds); | |
155 | jffs2_free_full_dirent(fd); | |
156 | } | |
157 | ||
158 | dbg_fsbuild("pass 2a complete\n"); | |
159 | dbg_fsbuild("freeing temporary data structures\n"); | |
160 | ||
161 | /* Finally, we can scan again and free the dirent structs */ | |
162 | for_each_inode(i, c, ic) { | |
163 | while(ic->scan_dents) { | |
164 | fd = ic->scan_dents; | |
165 | ic->scan_dents = fd->next; | |
166 | jffs2_free_full_dirent(fd); | |
167 | } | |
168 | ic->scan_dents = NULL; | |
169 | cond_resched(); | |
170 | } | |
171 | jffs2_build_xattr_subsystem(c); | |
172 | c->flags &= ~JFFS2_SB_FLAG_BUILDING; | |
173 | ||
174 | dbg_fsbuild("FS build complete\n"); | |
175 | ||
176 | /* Rotate the lists by some number to ensure wear levelling */ | |
177 | jffs2_rotate_lists(c); | |
178 | ||
179 | ret = 0; | |
180 | ||
181 | exit: | |
182 | if (ret) { | |
183 | for_each_inode(i, c, ic) { | |
184 | while(ic->scan_dents) { | |
185 | fd = ic->scan_dents; | |
186 | ic->scan_dents = fd->next; | |
187 | jffs2_free_full_dirent(fd); | |
188 | } | |
189 | } | |
190 | jffs2_clear_xattr_subsystem(c); | |
191 | } | |
192 | ||
193 | return ret; | |
194 | } | |
195 | ||
196 | static void jffs2_build_remove_unlinked_inode(struct jffs2_sb_info *c, | |
197 | struct jffs2_inode_cache *ic, | |
198 | struct jffs2_full_dirent **dead_fds) | |
199 | { | |
200 | struct jffs2_raw_node_ref *raw; | |
201 | struct jffs2_full_dirent *fd; | |
202 | ||
203 | dbg_fsbuild("removing ino #%u with nlink == zero.\n", ic->ino); | |
204 | ||
205 | raw = ic->nodes; | |
206 | while (raw != (void *)ic) { | |
207 | struct jffs2_raw_node_ref *next = raw->next_in_ino; | |
208 | dbg_fsbuild("obsoleting node at 0x%08x\n", ref_offset(raw)); | |
209 | jffs2_mark_node_obsolete(c, raw); | |
210 | raw = next; | |
211 | } | |
212 | ||
213 | if (ic->scan_dents) { | |
214 | int whinged = 0; | |
215 | dbg_fsbuild("inode #%u was a directory which may have children...\n", ic->ino); | |
216 | ||
217 | while(ic->scan_dents) { | |
218 | struct jffs2_inode_cache *child_ic; | |
219 | ||
220 | fd = ic->scan_dents; | |
221 | ic->scan_dents = fd->next; | |
222 | ||
223 | if (!fd->ino) { | |
224 | /* It's a deletion dirent. Ignore it */ | |
225 | dbg_fsbuild("child \"%s\" is a deletion dirent, skipping...\n", fd->name); | |
226 | jffs2_free_full_dirent(fd); | |
227 | continue; | |
228 | } | |
229 | if (!whinged) | |
230 | whinged = 1; | |
231 | ||
232 | dbg_fsbuild("removing child \"%s\", ino #%u\n", fd->name, fd->ino); | |
233 | ||
234 | child_ic = jffs2_get_ino_cache(c, fd->ino); | |
235 | if (!child_ic) { | |
236 | dbg_fsbuild("cannot remove child \"%s\", ino #%u, because it doesn't exist\n", | |
237 | fd->name, fd->ino); | |
238 | jffs2_free_full_dirent(fd); | |
239 | continue; | |
240 | } | |
241 | ||
242 | /* Reduce nlink of the child. If it's now zero, stick it on the | |
243 | dead_fds list to be cleaned up later. Else just free the fd */ | |
244 | ||
245 | if (fd->type == DT_DIR) | |
246 | child_ic->pino_nlink = 0; | |
247 | else | |
248 | child_ic->pino_nlink--; | |
249 | ||
250 | if (!child_ic->pino_nlink) { | |
251 | dbg_fsbuild("inode #%u (\"%s\") now has no links; adding to dead_fds list.\n", | |
252 | fd->ino, fd->name); | |
253 | fd->next = *dead_fds; | |
254 | *dead_fds = fd; | |
255 | } else { | |
256 | dbg_fsbuild("inode #%u (\"%s\") has now got nlink %d. Ignoring.\n", | |
257 | fd->ino, fd->name, child_ic->pino_nlink); | |
258 | jffs2_free_full_dirent(fd); | |
259 | } | |
260 | } | |
261 | } | |
262 | ||
263 | /* | |
264 | We don't delete the inocache from the hash list and free it yet. | |
265 | The erase code will do that, when all the nodes are completely gone. | |
266 | */ | |
267 | } | |
268 | ||
269 | static void jffs2_calc_trigger_levels(struct jffs2_sb_info *c) | |
270 | { | |
271 | uint32_t size; | |
272 | ||
273 | /* Deletion should almost _always_ be allowed. We're fairly | |
274 | buggered once we stop allowing people to delete stuff | |
275 | because there's not enough free space... */ | |
276 | c->resv_blocks_deletion = 2; | |
277 | ||
278 | /* Be conservative about how much space we need before we allow writes. | |
279 | On top of that which is required for deletia, require an extra 2% | |
280 | of the medium to be available, for overhead caused by nodes being | |
281 | split across blocks, etc. */ | |
282 | ||
283 | size = c->flash_size / 50; /* 2% of flash size */ | |
284 | size += c->nr_blocks * 100; /* And 100 bytes per eraseblock */ | |
285 | size += c->sector_size - 1; /* ... and round up */ | |
286 | ||
287 | c->resv_blocks_write = c->resv_blocks_deletion + (size / c->sector_size); | |
288 | ||
289 | /* When do we let the GC thread run in the background */ | |
290 | ||
291 | c->resv_blocks_gctrigger = c->resv_blocks_write + 1; | |
292 | ||
293 | /* When do we allow garbage collection to merge nodes to make | |
294 | long-term progress at the expense of short-term space exhaustion? */ | |
295 | c->resv_blocks_gcmerge = c->resv_blocks_deletion + 1; | |
296 | ||
297 | /* When do we allow garbage collection to eat from bad blocks rather | |
298 | than actually making progress? */ | |
299 | c->resv_blocks_gcbad = 0;//c->resv_blocks_deletion + 2; | |
300 | ||
301 | /* What number of 'very dirty' eraseblocks do we allow before we | |
302 | trigger the GC thread even if we don't _need_ the space. When we | |
303 | can't mark nodes obsolete on the medium, the old dirty nodes cause | |
304 | performance problems because we have to inspect and discard them. */ | |
305 | c->vdirty_blocks_gctrigger = c->resv_blocks_gctrigger; | |
306 | if (jffs2_can_mark_obsolete(c)) | |
307 | c->vdirty_blocks_gctrigger *= 10; | |
308 | ||
309 | /* If there's less than this amount of dirty space, don't bother | |
310 | trying to GC to make more space. It'll be a fruitless task */ | |
311 | c->nospc_dirty_size = c->sector_size + (c->flash_size / 100); | |
312 | ||
313 | dbg_fsbuild("trigger levels (size %d KiB, block size %d KiB, %d blocks)\n", | |
314 | c->flash_size / 1024, c->sector_size / 1024, c->nr_blocks); | |
315 | dbg_fsbuild("Blocks required to allow deletion: %d (%d KiB)\n", | |
316 | c->resv_blocks_deletion, c->resv_blocks_deletion*c->sector_size/1024); | |
317 | dbg_fsbuild("Blocks required to allow writes: %d (%d KiB)\n", | |
318 | c->resv_blocks_write, c->resv_blocks_write*c->sector_size/1024); | |
319 | dbg_fsbuild("Blocks required to quiesce GC thread: %d (%d KiB)\n", | |
320 | c->resv_blocks_gctrigger, c->resv_blocks_gctrigger*c->sector_size/1024); | |
321 | dbg_fsbuild("Blocks required to allow GC merges: %d (%d KiB)\n", | |
322 | c->resv_blocks_gcmerge, c->resv_blocks_gcmerge*c->sector_size/1024); | |
323 | dbg_fsbuild("Blocks required to GC bad blocks: %d (%d KiB)\n", | |
324 | c->resv_blocks_gcbad, c->resv_blocks_gcbad*c->sector_size/1024); | |
325 | dbg_fsbuild("Amount of dirty space required to GC: %d bytes\n", | |
326 | c->nospc_dirty_size); | |
327 | dbg_fsbuild("Very dirty blocks before GC triggered: %d\n", | |
328 | c->vdirty_blocks_gctrigger); | |
329 | } | |
330 | ||
331 | int jffs2_do_mount_fs(struct jffs2_sb_info *c) | |
332 | { | |
333 | int ret; | |
334 | int i; | |
335 | int size; | |
336 | ||
337 | c->free_size = c->flash_size; | |
338 | c->nr_blocks = c->flash_size / c->sector_size; | |
339 | size = sizeof(struct jffs2_eraseblock) * c->nr_blocks; | |
340 | #ifndef __ECOS | |
341 | if (jffs2_blocks_use_vmalloc(c)) | |
342 | c->blocks = vzalloc(size); | |
343 | else | |
344 | #endif | |
345 | c->blocks = kzalloc(size, GFP_KERNEL); | |
346 | if (!c->blocks) | |
347 | return -ENOMEM; | |
348 | ||
349 | for (i=0; i<c->nr_blocks; i++) { | |
350 | INIT_LIST_HEAD(&c->blocks[i].list); | |
351 | c->blocks[i].offset = i * c->sector_size; | |
352 | c->blocks[i].free_size = c->sector_size; | |
353 | } | |
354 | ||
355 | INIT_LIST_HEAD(&c->clean_list); | |
356 | INIT_LIST_HEAD(&c->very_dirty_list); | |
357 | INIT_LIST_HEAD(&c->dirty_list); | |
358 | INIT_LIST_HEAD(&c->erasable_list); | |
359 | INIT_LIST_HEAD(&c->erasing_list); | |
360 | INIT_LIST_HEAD(&c->erase_checking_list); | |
361 | INIT_LIST_HEAD(&c->erase_pending_list); | |
362 | INIT_LIST_HEAD(&c->erasable_pending_wbuf_list); | |
363 | INIT_LIST_HEAD(&c->erase_complete_list); | |
364 | INIT_LIST_HEAD(&c->free_list); | |
365 | INIT_LIST_HEAD(&c->bad_list); | |
366 | INIT_LIST_HEAD(&c->bad_used_list); | |
367 | c->highest_ino = 1; | |
368 | c->summary = NULL; | |
369 | ||
370 | ret = jffs2_sum_init(c); | |
371 | if (ret) | |
372 | goto out_free; | |
373 | ||
374 | if (jffs2_build_filesystem(c)) { | |
375 | dbg_fsbuild("build_fs failed\n"); | |
376 | jffs2_free_ino_caches(c); | |
377 | jffs2_free_raw_node_refs(c); | |
378 | ret = -EIO; | |
379 | goto out_free; | |
380 | } | |
381 | ||
382 | jffs2_calc_trigger_levels(c); | |
383 | ||
384 | return 0; | |
385 | ||
386 | out_free: | |
387 | kvfree(c->blocks); | |
388 | ||
389 | return ret; | |
390 | } |