* ubifs_get_free_space - return amount of free space.
* @c: UBIFS file-system description object
*
- * This function calculates and retuns amount of free space to report to
+ * This function calculates and returns amount of free space to report to
* user-space.
*/
long long ubifs_get_free_space(struct ubifs_info *c)
int lnum, offs, len, err = 0, uninitialized_var(last_level), child_cnt;
int first = 1, iip;
struct ubifs_debug_info *d = c->dbg;
- union ubifs_key lower_key, upper_key, l_key, u_key;
+ union ubifs_key uninitialized_var(lower_key), upper_key, l_key, u_key;
unsigned long long uninitialized_var(last_sqnum);
struct ubifs_idx_node *idx;
struct list_head list;
}
}
+const char *dbg_jhead(int jhead)
+{
+ switch (jhead) {
+ case GCHD:
+ return "0 (GC)";
+ case BASEHD:
+ return "1 (base)";
+ case DATAHD:
+ return "2 (data)";
+ default:
+ return "unknown journal head";
+ }
+}
+
static void dump_ch(const struct ubifs_ch *ch)
{
printk(KERN_DEBUG "\tmagic %#x\n", le32_to_cpu(ch->magic));
/* If we are in R/O mode, journal heads do not exist */
if (c->jheads)
for (i = 0; i < c->jhead_cnt; i++)
- printk(KERN_DEBUG "\tjhead %d\t LEB %d\n",
- c->jheads[i].wbuf.jhead, c->jheads[i].wbuf.lnum);
+ printk(KERN_DEBUG "\tjhead %s\t LEB %d\n",
+ dbg_jhead(c->jheads[i].wbuf.jhead),
+ c->jheads[i].wbuf.lnum);
for (rb = rb_first(&c->buds); rb; rb = rb_next(rb)) {
bud = rb_entry(rb, struct ubifs_bud, rb);
printk(KERN_DEBUG "\tbud LEB %d\n", bud->lnum);
void dbg_dump_lprop(const struct ubifs_info *c, const struct ubifs_lprops *lp)
{
- printk(KERN_DEBUG "LEB %d lprops: free %d, dirty %d (used %d), "
- "flags %#x\n", lp->lnum, lp->free, lp->dirty,
- c->leb_size - lp->free - lp->dirty, lp->flags);
+ int i, spc, dark = 0, dead = 0;
+ struct rb_node *rb;
+ struct ubifs_bud *bud;
+
+ spc = lp->free + lp->dirty;
+ if (spc < c->dead_wm)
+ dead = spc;
+ else
+ dark = ubifs_calc_dark(c, spc);
+
+ if (lp->flags & LPROPS_INDEX)
+ printk(KERN_DEBUG "LEB %-7d free %-8d dirty %-8d used %-8d "
+ "free + dirty %-8d flags %#x (", lp->lnum, lp->free,
+ lp->dirty, c->leb_size - spc, spc, lp->flags);
+ else
+ printk(KERN_DEBUG "LEB %-7d free %-8d dirty %-8d used %-8d "
+ "free + dirty %-8d dark %-4d dead %-4d nodes fit %-3d "
+ "flags %#-4x (", lp->lnum, lp->free, lp->dirty,
+ c->leb_size - spc, spc, dark, dead,
+ (int)(spc / UBIFS_MAX_NODE_SZ), lp->flags);
+
+ if (lp->flags & LPROPS_TAKEN) {
+ if (lp->flags & LPROPS_INDEX)
+ printk(KERN_CONT "index, taken");
+ else
+ printk(KERN_CONT "taken");
+ } else {
+ const char *s;
+
+ if (lp->flags & LPROPS_INDEX) {
+ switch (lp->flags & LPROPS_CAT_MASK) {
+ case LPROPS_DIRTY_IDX:
+ s = "dirty index";
+ break;
+ case LPROPS_FRDI_IDX:
+ s = "freeable index";
+ break;
+ default:
+ s = "index";
+ }
+ } else {
+ switch (lp->flags & LPROPS_CAT_MASK) {
+ case LPROPS_UNCAT:
+ s = "not categorized";
+ break;
+ case LPROPS_DIRTY:
+ s = "dirty";
+ break;
+ case LPROPS_FREE:
+ s = "free";
+ break;
+ case LPROPS_EMPTY:
+ s = "empty";
+ break;
+ case LPROPS_FREEABLE:
+ s = "freeable";
+ break;
+ default:
+ s = NULL;
+ break;
+ }
+ }
+ printk(KERN_CONT "%s", s);
+ }
+
+ for (rb = rb_first((struct rb_root *)&c->buds); rb; rb = rb_next(rb)) {
+ bud = rb_entry(rb, struct ubifs_bud, rb);
+ if (bud->lnum == lp->lnum) {
+ int head = 0;
+ for (i = 0; i < c->jhead_cnt; i++) {
+ if (lp->lnum == c->jheads[i].wbuf.lnum) {
+ printk(KERN_CONT ", jhead %s",
+ dbg_jhead(i));
+ head = 1;
+ }
+ }
+ if (!head)
+ printk(KERN_CONT ", bud of jhead %s",
+ dbg_jhead(bud->jhead));
+ }
+ }
+ if (lp->lnum == c->gc_lnum)
+ printk(KERN_CONT ", GC LEB");
+ printk(KERN_CONT ")\n");
}
void dbg_dump_lprops(struct ubifs_info *c)
printk(KERN_DEBUG "(pid %d) start dumping LEB %d\n",
current->pid, lnum);
- sleb = ubifs_scan(c, lnum, 0, c->dbg->buf);
+ sleb = ubifs_scan(c, lnum, 0, c->dbg->buf, 0);
if (IS_ERR(sleb)) {
ubifs_err("scan error %d", (int)PTR_ERR(sleb));
return;
ubifs_msg("saved lprops statistics dump");
dbg_dump_lstats(&d->saved_lst);
ubifs_get_lp_stats(c, &lst);
+
ubifs_msg("current lprops statistics dump");
- dbg_dump_lstats(&d->saved_lst);
+ dbg_dump_lstats(&lst);
+
spin_lock(&c->space_lock);
dbg_dump_budg(c);
spin_unlock(&c->space_lock);
/* Dump functions */
const char *dbg_ntype(int type);
const char *dbg_cstate(int cmt_state);
+const char *dbg_jhead(int jhead);
const char *dbg_get_key_dump(const struct ubifs_info *c,
const union ubifs_key *key);
void dbg_dump_inode(const struct ubifs_info *c, const struct inode *inode);
int dbg_check_lprops(struct ubifs_info *c);
int dbg_check_lpt_nodes(struct ubifs_info *c, struct ubifs_cnode *cnode,
int row, int col);
+int dbg_check_inode_size(struct ubifs_info *c, const struct inode *inode,
+ loff_t size);
/* Force the use of in-the-gaps method for testing */
#define dbg_ntype(type) ""
#define dbg_cstate(cmt_state) ""
+#define dbg_jhead(jhead) ""
#define dbg_get_key_dump(c, key) ({})
#define dbg_dump_inode(c, inode) ({})
#define dbg_dump_node(c, node) ({})
#define dbg_check_heap(c, heap, cat, add_pos) ({})
#define dbg_check_lprops(c) 0
#define dbg_check_lpt_nodes(c, cnode, row, col) 0
+#define dbg_check_inode_size(c, inode, size) 0
#define dbg_force_in_the_gaps_enabled 0
#define dbg_force_in_the_gaps() 0
#define dbg_failure_mode 0
*/
/*
- * This file implements VFS file and inode operations of regular files, device
+ * This file implements VFS file and inode operations for regular files, device
* nodes and symlinks as well as address space operations.
*
- * UBIFS uses 2 page flags: PG_private and PG_checked. PG_private is set if the
- * page is dirty and is used for budgeting purposes - dirty pages should not be
- * budgeted. The PG_checked flag is set if full budgeting is required for the
- * page e.g., when it corresponds to a file hole or it is just beyond the file
- * size. The budgeting is done in 'ubifs_write_begin()', because it is OK to
- * fail in this function, and the budget is released in 'ubifs_write_end()'. So
- * the PG_private and PG_checked flags carry the information about how the page
- * was budgeted, to make it possible to release the budget properly.
+ * UBIFS uses 2 page flags: @PG_private and @PG_checked. @PG_private is set if
+ * the page is dirty and is used for optimization purposes - dirty pages are
+ * not budgeted so the flag shows that 'ubifs_write_end()' should not release
+ * the budget for this page. The @PG_checked flag is set if full budgeting is
+ * required for the page e.g., when it corresponds to a file hole or it is
+ * beyond the file size. The budgeting is done in 'ubifs_write_begin()', because
+ * it is OK to fail in this function, and the budget is released in
+ * 'ubifs_write_end()'. So the @PG_private and @PG_checked flags carry
+ * information about how the page was budgeted, to make it possible to release
+ * the budget properly.
*
- * A thing to keep in mind: inode's 'i_mutex' is locked in most VFS operations
- * we implement. However, this is not true for '->writepage()', which might be
- * called with 'i_mutex' unlocked. For example, when pdflush is performing
- * write-back, it calls 'writepage()' with unlocked 'i_mutex', although the
- * inode has 'I_LOCK' flag in this case. At "normal" work-paths 'i_mutex' is
- * locked in '->writepage', e.g. in "sys_write -> alloc_pages -> direct reclaim
- * path'. So, in '->writepage()' we are only guaranteed that the page is
- * locked.
+ * A thing to keep in mind: inode @i_mutex is locked in most VFS operations we
+ * implement. However, this is not true for 'ubifs_writepage()', which may be
+ * called with @i_mutex unlocked. For example, when pdflush is doing background
+ * write-back, it calls 'ubifs_writepage()' with unlocked @i_mutex. At "normal"
+ * work-paths the @i_mutex is locked in 'ubifs_writepage()', e.g. in the
+ * "sys_write -> alloc_pages -> direct reclaim path". So, in 'ubifs_writepage()'
+ * we are only guaranteed that the page is locked.
*
- * Similarly, 'i_mutex' does not have to be locked in readpage(), e.g.,
- * readahead path does not have it locked ("sys_read -> generic_file_aio_read
- * -> ondemand_readahead -> readpage"). In case of readahead, 'I_LOCK' flag is
- * not set as well. However, UBIFS disables readahead.
- *
- * This, for example means that there might be 2 concurrent '->writepage()'
- * calls for the same inode, but different inode dirty pages.
+ * Similarly, @i_mutex is not always locked in 'ubifs_readpage()', e.g., the
+ * read-ahead path does not lock it ("sys_read -> generic_file_aio_read ->
+ * ondemand_readahead -> readpage"). In case of readahead, @I_LOCK flag is not
+ * set as well. However, UBIFS disables readahead.
*/
#include "ubifs.h"
/*
* We change whole page so no need to load it. But we
* have to set the @PG_checked flag to make the further
- * code the page is new. This might be not true, but it
- * is better to budget more that to read the page from
- * the media.
+ * code know that the page is new. This might be not
+ * true, but it is better to budget more than to read
+ * the page from the media.
*/
SetPageChecked(page);
skipped_read = 1;
}
/*
- * Whee, we aquired budgeting quickly - without involving
- * garbage-collection, committing or forceing write-back. We return
+ * Whee, we acquired budgeting quickly - without involving
+ * garbage-collection, committing or forcing write-back. We return
* with @ui->ui_mutex locked if we are appending pages, and unlocked
* otherwise. This is an optimization (slightly hacky though).
*/
/*
* Return 0 to force VFS to repeat the whole operation, or the
- * error code if 'do_readpage()' failes.
+ * error code if 'do_readpage()' fails.
*/
copied = do_readpage(page);
goto out;
ui->ui_size = inode->i_size;
/* Truncation changes inode [mc]time */
inode->i_mtime = inode->i_ctime = ubifs_current_time(inode);
- /* The other attributes may be changed at the same time as well */
+ /* Other attributes may be changed at the same time as well */
do_attr_changes(inode, attr);
-
err = ubifs_jnl_truncate(c, inode, old_size, new_size);
mutex_unlock(&ui->ui_mutex);
+
out_budg:
if (budgeted)
ubifs_release_budget(c, &req);
* We scan the entire LEB even though we only really need to scan up to
* (c->leb_size - lp->free).
*/
- sleb = ubifs_scan(c, lnum, 0, c->sbuf);
+ sleb = ubifs_scan(c, lnum, 0, c->sbuf, 0);
if (IS_ERR(sleb))
return PTR_ERR(sleb);
{
struct ubifs_wbuf *wbuf = container_of(timer, struct ubifs_wbuf, timer);
- dbg_io("jhead %d", wbuf->jhead);
+ dbg_io("jhead %s", dbg_jhead(wbuf->jhead));
wbuf->need_sync = 1;
wbuf->c->need_wbuf_sync = 1;
ubifs_wake_up_bgt(wbuf->c);
if (wbuf->no_timer)
return;
- dbg_io("set timer for jhead %d, %llu-%llu millisecs", wbuf->jhead,
+ dbg_io("set timer for jhead %s, %llu-%llu millisecs",
+ dbg_jhead(wbuf->jhead),
div_u64(ktime_to_ns(wbuf->softlimit), USEC_PER_SEC),
div_u64(ktime_to_ns(wbuf->softlimit) + wbuf->delta,
USEC_PER_SEC));
/* Write-buffer is empty or not seeked */
return 0;
- dbg_io("LEB %d:%d, %d bytes, jhead %d",
- wbuf->lnum, wbuf->offs, wbuf->used, wbuf->jhead);
+ dbg_io("LEB %d:%d, %d bytes, jhead %s",
+ wbuf->lnum, wbuf->offs, wbuf->used, dbg_jhead(wbuf->jhead));
ubifs_assert(!(c->vfs_sb->s_flags & MS_RDONLY));
ubifs_assert(!(wbuf->avail & 7));
ubifs_assert(wbuf->offs + c->min_io_size <= c->leb_size);
{
const struct ubifs_info *c = wbuf->c;
- dbg_io("LEB %d:%d, jhead %d", lnum, offs, wbuf->jhead);
+ dbg_io("LEB %d:%d, jhead %s", lnum, offs, dbg_jhead(wbuf->jhead));
ubifs_assert(lnum >= 0 && lnum < c->leb_cnt);
ubifs_assert(offs >= 0 && offs <= c->leb_size);
ubifs_assert(offs % c->min_io_size == 0 && !(offs & 7));
struct ubifs_info *c = wbuf->c;
int err, written, n, aligned_len = ALIGN(len, 8), offs;
- dbg_io("%d bytes (%s) to jhead %d wbuf at LEB %d:%d", len,
- dbg_ntype(((struct ubifs_ch *)buf)->node_type), wbuf->jhead,
- wbuf->lnum, wbuf->offs + wbuf->used);
+ dbg_io("%d bytes (%s) to jhead %s wbuf at LEB %d:%d", len,
+ dbg_ntype(((struct ubifs_ch *)buf)->node_type),
+ dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs + wbuf->used);
ubifs_assert(len > 0 && wbuf->lnum >= 0 && wbuf->lnum < c->leb_cnt);
ubifs_assert(wbuf->offs >= 0 && wbuf->offs % c->min_io_size == 0);
ubifs_assert(!(wbuf->offs & 7) && wbuf->offs <= c->leb_size);
memcpy(wbuf->buf + wbuf->used, buf, len);
if (aligned_len == wbuf->avail) {
- dbg_io("flush jhead %d wbuf to LEB %d:%d",
- wbuf->jhead, wbuf->lnum, wbuf->offs);
+ dbg_io("flush jhead %s wbuf to LEB %d:%d",
+ dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs);
err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf,
wbuf->offs, c->min_io_size,
wbuf->dtype);
* minimal I/O unit. We have to fill and flush write-buffer and switch
* to the next min. I/O unit.
*/
- dbg_io("flush jhead %d wbuf to LEB %d:%d",
- wbuf->jhead, wbuf->lnum, wbuf->offs);
+ dbg_io("flush jhead %s wbuf to LEB %d:%d",
+ dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs);
memcpy(wbuf->buf + wbuf->used, buf, wbuf->avail);
err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs,
c->min_io_size, wbuf->dtype);
int err, rlen, overlap;
struct ubifs_ch *ch = buf;
- dbg_io("LEB %d:%d, %s, length %d, jhead %d", lnum, offs,
- dbg_ntype(type), len, wbuf->jhead);
+ dbg_io("LEB %d:%d, %s, length %d, jhead %s", lnum, offs,
+ dbg_ntype(type), len, dbg_jhead(wbuf->jhead));
ubifs_assert(wbuf && lnum >= 0 && lnum < c->leb_cnt && offs >= 0);
ubifs_assert(!(offs & 7) && offs < c->leb_size);
ubifs_assert(type >= 0 && type < UBIFS_NODE_TYPES_CNT);
* some. But the write-buffer mutex has to be unlocked because
* GC also takes it.
*/
- dbg_jnl("no free space jhead %d, run GC", jhead);
+ dbg_jnl("no free space in jhead %s, run GC", dbg_jhead(jhead));
mutex_unlock(&wbuf->io_mutex);
lnum = ubifs_garbage_collect(c, 0);
* because we dropped @wbuf->io_mutex, so try once
* again.
*/
- dbg_jnl("GC couldn't make a free LEB for jhead %d", jhead);
+ dbg_jnl("GC couldn't make a free LEB for jhead %s",
+ dbg_jhead(jhead));
if (retries++ < 2) {
dbg_jnl("retry (%d)", retries);
goto again;
}
mutex_lock_nested(&wbuf->io_mutex, wbuf->jhead);
- dbg_jnl("got LEB %d for jhead %d", lnum, jhead);
+ dbg_jnl("got LEB %d for jhead %s", lnum, dbg_jhead(jhead));
avail = c->leb_size - wbuf->offs - wbuf->used;
if (wbuf->lnum != -1 && avail >= len) {
*lnum = c->jheads[jhead].wbuf.lnum;
*offs = c->jheads[jhead].wbuf.offs + c->jheads[jhead].wbuf.used;
- dbg_jnl("jhead %d, LEB %d:%d, len %d", jhead, *lnum, *offs, len);
+ dbg_jnl("jhead %s, LEB %d:%d, len %d",
+ dbg_jhead(jhead), *lnum, *offs, len);
ubifs_prepare_node(c, node, len, 0);
return ubifs_wbuf_write_nolock(wbuf, node, len);
*lnum = c->jheads[jhead].wbuf.lnum;
*offs = c->jheads[jhead].wbuf.offs + c->jheads[jhead].wbuf.used;
- dbg_jnl("jhead %d, LEB %d:%d, len %d", jhead, *lnum, *offs, len);
+ dbg_jnl("jhead %s, LEB %d:%d, len %d",
+ dbg_jhead(jhead), *lnum, *offs, len);
err = ubifs_wbuf_write_nolock(wbuf, buf, len);
if (err)
}
/**
- * xent_key_init_hash - initialize extended attribute entry key without
- * re-calculating hash function.
- * @c: UBIFS file-system description object
- * @key: key to initialize
- * @inum: host inode number
- * @hash: extended attribute entry name hash
- */
-static inline void xent_key_init_hash(const struct ubifs_info *c,
- union ubifs_key *key, ino_t inum,
- uint32_t hash)
-{
- ubifs_assert(!(hash & ~UBIFS_S_KEY_HASH_MASK));
- key->u32[0] = inum;
- key->u32[1] = hash | (UBIFS_XENT_KEY << UBIFS_S_KEY_HASH_BITS);
-}
-
-/**
* xent_key_init_flash - initialize on-flash extended attribute entry key.
* @c: UBIFS file-system description object
* @k: key to initialize
}
/**
- * data_key_init_flash - initialize on-flash data key.
+ * highest_data_key - get the highest possible data key for an inode.
* @c: UBIFS file-system description object
- * @k: key to initialize
+ * @key: key to initialize
* @inum: inode number
- * @block: block number
*/
-static inline void data_key_init_flash(const struct ubifs_info *c, void *k,
- ino_t inum, unsigned int block)
+static inline void highest_data_key(const struct ubifs_info *c,
+ union ubifs_key *key, ino_t inum)
{
- union ubifs_key *key = k;
-
- ubifs_assert(!(block & ~UBIFS_S_KEY_BLOCK_MASK));
- key->j32[0] = cpu_to_le32(inum);
- key->j32[1] = cpu_to_le32(block |
- (UBIFS_DATA_KEY << UBIFS_S_KEY_BLOCK_BITS));
- memset(k + 8, 0, UBIFS_MAX_KEY_LEN - 8);
+ data_key_init(c, key, inum, UBIFS_S_KEY_BLOCK_MASK);
}
/**
return 0;
}
}
+
#endif /* !__UBIFS_KEY_H__ */
*/
c->bud_bytes += c->leb_size - bud->start;
- dbg_log("LEB %d:%d, jhead %d, bud_bytes %lld", bud->lnum,
- bud->start, bud->jhead, c->bud_bytes);
+ dbg_log("LEB %d:%d, jhead %s, bud_bytes %lld", bud->lnum,
+ bud->start, dbg_jhead(bud->jhead), c->bud_bytes);
spin_unlock(&c->buds_lock);
}
* heads (non-closed buds).
*/
c->cmt_bud_bytes += wbuf->offs - bud->start;
- dbg_log("preserve %d:%d, jhead %d, bud bytes %d, "
+ dbg_log("preserve %d:%d, jhead %s, bud bytes %d, "
"cmt_bud_bytes %lld", bud->lnum, bud->start,
- bud->jhead, wbuf->offs - bud->start,
+ dbg_jhead(bud->jhead), wbuf->offs - bud->start,
c->cmt_bud_bytes);
bud->start = wbuf->offs;
} else {
c->cmt_bud_bytes += c->leb_size - bud->start;
- dbg_log("remove %d:%d, jhead %d, bud bytes %d, "
+ dbg_log("remove %d:%d, jhead %s, bud bytes %d, "
"cmt_bud_bytes %lld", bud->lnum, bud->start,
- bud->jhead, c->leb_size - bud->start,
+ dbg_jhead(bud->jhead), c->leb_size - bud->start,
c->cmt_bud_bytes);
rb_erase(p1, &c->buds);
/*
if (lnum == -1 || offs == c->leb_size)
continue;
- dbg_log("add ref to LEB %d:%d for jhead %d", lnum, offs, i);
+ dbg_log("add ref to LEB %d:%d for jhead %s",
+ lnum, offs, dbg_jhead(i));
ref = buf + len;
ref->ch.node_type = UBIFS_REF_NODE;
ref->lnum = cpu_to_le32(lnum);
lnum = c->ltail_lnum;
write_lnum = lnum;
while (1) {
- sleb = ubifs_scan(c, lnum, 0, c->sbuf);
+ sleb = ubifs_scan(c, lnum, 0, c->sbuf, 0);
if (IS_ERR(sleb)) {
err = PTR_ERR(sleb);
goto out_free;
case LPROPS_FREE:
if (add_to_lpt_heap(c, lprops, cat))
break;
- /* No more room on heap so make it uncategorized */
+ /* No more room on heap so make it un-categorized */
cat = LPROPS_UNCAT;
/* Fall through */
case LPROPS_UNCAT:
* @lprops: LEB properties
*
* A LEB may have fallen off of the bottom of a heap, and ended up as
- * uncategorized even though it has enough space for us now. If that is the case
- * this function will put the LEB back onto a heap.
+ * un-categorized even though it has enough space for us now. If that is the
+ * case this function will put the LEB back onto a heap.
*/
void ubifs_ensure_cat(struct ubifs_info *c, struct ubifs_lprops *lprops)
{
/**
* change_category - change LEB properties category.
* @c: UBIFS file-system description object
- * @lprops: LEB properties to recategorize
+ * @lprops: LEB properties to re-categorize
*
* LEB properties are categorized to enable fast find operations. When the LEB
- * properties change they must be recategorized.
+ * properties change they must be re-categorized.
*/
static void change_category(struct ubifs_info *c, struct ubifs_lprops *lprops)
{
}
/**
- * calc_dark - calculate LEB dark space size.
+ * ubifs_calc_dark - calculate LEB dark space size.
* @c: the UBIFS file-system description object
* @spc: amount of free and dirty space in the LEB
*
- * This function calculates amount of dark space in an LEB which has @spc bytes
- * of free and dirty space. Returns the calculations result.
+ * This function calculates and returns amount of dark space in an LEB which
+ * has @spc bytes of free and dirty space.
*
- * Dark space is the space which is not always usable - it depends on which
- * nodes are written in which order. E.g., if an LEB has only 512 free bytes,
- * it is dark space, because it cannot fit a large data node. So UBIFS cannot
- * count on this LEB and treat these 512 bytes as usable because it is not true
- * if, for example, only big chunks of uncompressible data will be written to
- * the FS.
+ * UBIFS is trying to account the space which might not be usable, and this
+ * space is called "dark space". For example, if an LEB has only %512 free
+ * bytes, it is dark space, because it cannot fit a large data node.
*/
-static int calc_dark(struct ubifs_info *c, int spc)
+int ubifs_calc_dark(const struct ubifs_info *c, int spc)
{
ubifs_assert(!(spc & 7));
* @free: new free space amount
* @dirty: new dirty space amount
* @flags: new flags
- * @idx_gc_cnt: change to the count of idx_gc list
+ * @idx_gc_cnt: change to the count of @idx_gc list
*
* This function changes LEB properties (@free, @dirty or @flag). However, the
* property which has the %LPROPS_NC value is not changed. Returns a pointer to
{
/*
* This is the only function that is allowed to change lprops, so we
- * discard the const qualifier.
+ * discard the "const" qualifier.
*/
struct ubifs_lprops *lprops = (struct ubifs_lprops *)lp;
if (old_spc < c->dead_wm)
c->lst.total_dead -= old_spc;
else
- c->lst.total_dark -= calc_dark(c, old_spc);
+ c->lst.total_dark -= ubifs_calc_dark(c, old_spc);
c->lst.total_used -= c->leb_size - old_spc;
}
if (new_spc < c->dead_wm)
c->lst.total_dead += new_spc;
else
- c->lst.total_dark += calc_dark(c, new_spc);
+ c->lst.total_dark += ubifs_calc_dark(c, new_spc);
c->lst.total_used += c->leb_size - new_spc;
}
}
}
- sleb = ubifs_scan(c, lnum, 0, c->dbg->buf);
+ sleb = ubifs_scan(c, lnum, 0, c->dbg->buf, 0);
if (IS_ERR(sleb)) {
/*
* After an unclean unmount, empty and freeable LEBs
"- continuing checking");
lst->empty_lebs += 1;
lst->total_free += c->leb_size;
- lst->total_dark += calc_dark(c, c->leb_size);
+ lst->total_dark += ubifs_calc_dark(c, c->leb_size);
return LPT_SCAN_CONTINUE;
}
"- continuing checking");
lst->total_free += lp->free;
lst->total_dirty += lp->dirty;
- lst->total_dark += calc_dark(c, c->leb_size);
+ lst->total_dark += ubifs_calc_dark(c, c->leb_size);
return LPT_SCAN_CONTINUE;
}
data->err = PTR_ERR(sleb);
if (spc < c->dead_wm)
lst->total_dead += spc;
else
- lst->total_dark += calc_dark(c, spc);
+ lst->total_dark += ubifs_calc_dark(c, spc);
}
ubifs_scan_destroy(sleb);
* @c: UBIFS file-system description object
*
* This function scans the master node LEBs and search for the latest master
- * node. Returns zero in case of success and a negative error code in case of
+ * node. Returns zero in case of success, %-EUCLEAN if there master area is
+ * corrupted and requires recovery, and a negative error code in case of
* failure.
*/
static int scan_for_master(struct ubifs_info *c)
lnum = UBIFS_MST_LNUM;
- sleb = ubifs_scan(c, lnum, 0, c->sbuf);
+ sleb = ubifs_scan(c, lnum, 0, c->sbuf, 1);
if (IS_ERR(sleb))
return PTR_ERR(sleb);
nodes_cnt = sleb->nodes_cnt;
snod = list_entry(sleb->nodes.prev, struct ubifs_scan_node,
list);
if (snod->type != UBIFS_MST_NODE)
- goto out;
+ goto out_dump;
memcpy(c->mst_node, snod->node, snod->len);
offs = snod->offs;
}
lnum += 1;
- sleb = ubifs_scan(c, lnum, 0, c->sbuf);
+ sleb = ubifs_scan(c, lnum, 0, c->sbuf, 1);
if (IS_ERR(sleb))
return PTR_ERR(sleb);
if (sleb->nodes_cnt != nodes_cnt)
goto out;
snod = list_entry(sleb->nodes.prev, struct ubifs_scan_node, list);
if (snod->type != UBIFS_MST_NODE)
- goto out;
+ goto out_dump;
if (snod->offs != offs)
goto out;
if (memcmp((void *)c->mst_node + UBIFS_CH_SZ,
out:
ubifs_scan_destroy(sleb);
+ return -EUCLEAN;
+
+out_dump:
+ ubifs_err("unexpected node type %d master LEB %d:%d",
+ snod->type, lnum, snod->offs);
+ ubifs_scan_destroy(sleb);
return -EINVAL;
}
err = scan_for_master(c);
if (err) {
- err = ubifs_recover_master_node(c);
+ if (err == -EUCLEAN)
+ err = ubifs_recover_master_node(c);
if (err)
/*
* Note, we do not free 'c->mst_node' here because the
struct ubifs_scan_leb *sleb;
dbg_rcvry("LEB %d", lnum);
- sleb = ubifs_scan(c, lnum, 0, c->sbuf);
+ sleb = ubifs_scan(c, lnum, 0, c->sbuf, 1);
if (IS_ERR(sleb)) {
- sleb = ubifs_recover_leb(c, lnum, 0, c->sbuf, 0);
+ if (PTR_ERR(sleb) == -EUCLEAN)
+ sleb = ubifs_recover_leb(c, lnum, 0, c->sbuf, 0);
if (IS_ERR(sleb)) {
err = PTR_ERR(sleb);
break;
for (lnum = c->orph_first; lnum <= c->orph_last; lnum++) {
struct ubifs_scan_leb *sleb;
- sleb = ubifs_scan(c, lnum, 0, c->dbg->buf);
+ sleb = ubifs_scan(c, lnum, 0, c->dbg->buf, 0);
if (IS_ERR(sleb)) {
err = PTR_ERR(sleb);
break;
mst = mst2;
}
- dbg_rcvry("recovered master node from LEB %d",
+ ubifs_msg("recovered master node from LEB %d",
(mst == mst1 ? UBIFS_MST_LNUM : UBIFS_MST_LNUM + 1));
memcpy(c->mst_node, mst, UBIFS_MST_NODE_SZ);
* We can only recover at the end of the log, so check that the
* next log LEB is empty or out of date.
*/
- sleb = ubifs_scan(c, next_lnum, 0, sbuf);
+ sleb = ubifs_scan(c, next_lnum, 0, sbuf, 0);
if (IS_ERR(sleb))
return sleb;
if (sleb->nodes_cnt) {
if (c->need_recovery)
sleb = ubifs_recover_leb(c, lnum, offs, c->sbuf, jhead != GCHD);
else
- sleb = ubifs_scan(c, lnum, offs, c->sbuf);
+ sleb = ubifs_scan(c, lnum, offs, c->sbuf, 0);
if (IS_ERR(sleb))
return PTR_ERR(sleb);
const struct ubifs_cs_node *node;
dbg_mnt("replay log LEB %d:%d", lnum, offs);
- sleb = ubifs_scan(c, lnum, offs, sbuf);
- if (IS_ERR(sleb) ) {
+ sleb = ubifs_scan(c, lnum, offs, sbuf, c->need_recovery);
+ if (IS_ERR(sleb)) {
if (PTR_ERR(sleb) != -EUCLEAN || !c->need_recovery)
return PTR_ERR(sleb);
sleb = ubifs_recover_log_leb(c, lnum, offs, sbuf);
/* Make the node pads to 8-byte boundary */
if ((node_len + pad_len) & 7) {
- if (!quiet) {
+ if (!quiet)
dbg_err("bad padding length %d - %d",
offs, offs + node_len + pad_len);
- }
return SCANNED_A_BAD_PAD_NODE;
}
* @c: UBIFS file-system description object
* @lnum: logical eraseblock number
* @offs: offset to start at (usually zero)
- * @sbuf: scan buffer (must be c->leb_size)
+ * @sbuf: scan buffer (must be of @c->leb_size bytes in size)
+ * @quiet: print no messages
*
* This function scans LEB number @lnum and returns complete information about
* its contents. Returns the scaned information in case of success and,
* %-EUCLEAN if the LEB neads recovery, and other negative error codes in case
* of failure.
+ *
+ * If @quiet is non-zero, this function does not print large and scary
+ * error messages and flash dumps in case of errors.
*/
struct ubifs_scan_leb *ubifs_scan(const struct ubifs_info *c, int lnum,
- int offs, void *sbuf)
+ int offs, void *sbuf, int quiet)
{
void *buf = sbuf + offs;
int err, len = c->leb_size - offs;
cond_resched();
- ret = ubifs_scan_a_node(c, buf, len, lnum, offs, 0);
+ ret = ubifs_scan_a_node(c, buf, len, lnum, offs, quiet);
if (ret > 0) {
/* Padding bytes or a valid padding node */
offs += ret;
}
if (offs % c->min_io_size) {
- ubifs_err("empty space starts at non-aligned offset %d", offs);
+ if (!quiet)
+ ubifs_err("empty space starts at non-aligned offset %d",
+ offs);
goto corrupted;;
}
break;
for (; len; offs++, buf++, len--)
if (*(uint8_t *)buf != 0xff) {
- ubifs_err("corrupt empty space at LEB %d:%d",
- lnum, offs);
+ if (!quiet)
+ ubifs_err("corrupt empty space at LEB %d:%d",
+ lnum, offs);
goto corrupted;
}
return sleb;
corrupted:
- ubifs_scanned_corruption(c, lnum, offs, buf);
+ if (!quiet) {
+ ubifs_scanned_corruption(c, lnum, offs, buf);
+ ubifs_err("LEB %d scanning failed", lnum);
+ }
err = -EUCLEAN;
+ ubifs_scan_destroy(sleb);
+ return ERR_PTR(err);
+
error:
- ubifs_err("LEB %d scanning failed", lnum);
+ ubifs_err("LEB %d scanning failed, error %d", lnum, err);
ubifs_scan_destroy(sleb);
return ERR_PTR(err);
}
#include <linux/mount.h>
#include <linux/math64.h>
#include <linux/writeback.h>
-#include <linux/smp_lock.h>
#include "ubifs.h"
/*
if (err)
ubifs_err("can't write inode %lu, error %d",
inode->i_ino, err);
+ else
+ err = dbg_check_inode_size(c, inode, ui->ui_size);
}
ui->dirty = 0;
return 0;
/*
- * VFS calls '->sync_fs()' before synchronizing all dirty inodes and
- * pages, so synchronize them first, then commit the journal. Strictly
- * speaking, it is not necessary to commit the journal here,
- * synchronizing write-buffers would be enough. But committing makes
- * UBIFS free space predictions much more accurate, so we want to let
- * the user be able to get more accurate results of 'statfs()' after
- * they synchronize the file system.
- */
- sync_inodes_sb(sb);
-
- /*
* Synchronize write buffers, because 'ubifs_run_commit()' does not
* do this if it waits for an already running commit.
*/
return err;
}
+ /*
+ * Strictly speaking, it is not necessary to commit the journal here,
+ * synchronizing write-buffers would be enough. But committing makes
+ * UBIFS free space predictions much more accurate, so we want to let
+ * the user be able to get more accurate results of 'statfs()' after
+ * they synchronize the file system.
+ */
err = ubifs_run_commit(c);
if (err)
return err;
ubifs_msg("un-mount UBI device %d, volume %d", c->vi.ubi_num,
c->vi.vol_id);
- lock_kernel();
-
/*
* The following asserts are only valid if there has not been a failure
* of the media. For example, there will be dirty inodes if we failed
ubi_close_volume(c->ubi);
mutex_unlock(&c->umount_mutex);
kfree(c);
-
- unlock_kernel();
}
static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data)
return err;
}
- lock_kernel();
if ((sb->s_flags & MS_RDONLY) && !(*flags & MS_RDONLY)) {
if (c->ro_media) {
ubifs_msg("cannot re-mount due to prior errors");
- unlock_kernel();
return -EROFS;
}
err = ubifs_remount_rw(c);
- if (err) {
- unlock_kernel();
+ if (err)
return err;
- }
} else if (!(sb->s_flags & MS_RDONLY) && (*flags & MS_RDONLY)) {
if (c->ro_media) {
ubifs_msg("cannot re-mount due to prior errors");
- unlock_kernel();
return -EROFS;
}
ubifs_remount_ro(c);
}
ubifs_assert(c->lst.taken_empty_lebs > 0);
- unlock_kernel();
return 0;
}
* o exact match, i.e. the found zero-level znode contains key @key, then %1
* is returned and slot number of the matched branch is stored in @n;
* o not exact match, which means that zero-level znode does not contain
- * @key, then %0 is returned and slot number of the closed branch is stored
- * in @n;
+ * @key, then %0 is returned and slot number of the closest branch is stored
+ * in @n;
* o @key is so small that it is even less than the lowest key of the
* leftmost zero-level node, then %0 is returned and %0 is stored in @n.
*
* @lnum: LEB number is returned here
* @offs: offset is returned here
*
- * This function look up and reads node with key @key. The caller has to make
+ * This function looks up and reads node with key @key. The caller has to make
* sure the @node buffer is large enough to fit the node. Returns zero in case
* of success, %-ENOENT if the node was not found, and a negative error code in
* case of failure. The node location can be returned in @lnum and @offs.
mutex_unlock(&c->tnc_mutex);
return err;
}
+
+#ifdef CONFIG_UBIFS_FS_DEBUG
+
+/**
+ * dbg_check_inode_size - check if inode size is correct.
+ * @c: UBIFS file-system description object
+ * @inum: inode number
+ * @size: inode size
+ *
+ * This function makes sure that the inode size (@size) is correct and it does
+ * not have any pages beyond @size. Returns zero if the inode is OK, %-EINVAL
+ * if it has a data page beyond @size, and other negative error code in case of
+ * other errors.
+ */
+int dbg_check_inode_size(struct ubifs_info *c, const struct inode *inode,
+ loff_t size)
+{
+ int err, n;
+ union ubifs_key from_key, to_key, *key;
+ struct ubifs_znode *znode;
+ unsigned int block;
+
+ if (!S_ISREG(inode->i_mode))
+ return 0;
+ if (!(ubifs_chk_flags & UBIFS_CHK_GEN))
+ return 0;
+
+ block = (size + UBIFS_BLOCK_SIZE - 1) >> UBIFS_BLOCK_SHIFT;
+ data_key_init(c, &from_key, inode->i_ino, block);
+ highest_data_key(c, &to_key, inode->i_ino);
+
+ mutex_lock(&c->tnc_mutex);
+ err = ubifs_lookup_level0(c, &from_key, &znode, &n);
+ if (err < 0)
+ goto out_unlock;
+
+ if (err) {
+ err = -EINVAL;
+ key = &from_key;
+ goto out_dump;
+ }
+
+ err = tnc_next(c, &znode, &n);
+ if (err == -ENOENT) {
+ err = 0;
+ goto out_unlock;
+ }
+ if (err < 0)
+ goto out_unlock;
+
+ ubifs_assert(err == 0);
+ key = &znode->zbranch[n].key;
+ if (!key_in_range(c, key, &from_key, &to_key))
+ goto out_unlock;
+
+out_dump:
+ block = key_block(c, key);
+ ubifs_err("inode %lu has size %lld, but there are data at offset %lld "
+ "(data key %s)", (unsigned long)inode->i_ino, size,
+ ((loff_t)block) << UBIFS_BLOCK_SHIFT, DBGKEY(key));
+ dbg_dump_inode(c, inode);
+ dbg_dump_stack();
+ err = -EINVAL;
+
+out_unlock:
+ mutex_unlock(&c->tnc_mutex);
+ return err;
+}
+
+#endif /* CONFIG_UBIFS_FS_DEBUG */
* it is more comprehensive and less efficient than is needed for this
* purpose.
*/
- sleb = ubifs_scan(c, lnum, 0, c->ileb_buf);
+ sleb = ubifs_scan(c, lnum, 0, c->ileb_buf, 0);
c->ileb_len = 0;
if (IS_ERR(sleb))
return PTR_ERR(sleb);
/* The key is always at the same position in all keyed nodes */
#define UBIFS_KEY_OFFSET offsetof(struct ubifs_ino_node, key)
+/* Garbage collector journal head number */
+#define UBIFS_GC_HEAD 0
+/* Base journal head number */
+#define UBIFS_BASE_HEAD 1
+/* Data journal head number */
+#define UBIFS_DATA_HEAD 2
+
/*
* LEB Properties Tree node types.
*
/* Number of non-data journal heads */
#define NONDATA_JHEADS_CNT 2
-/* Garbage collector head */
-#define GCHD 0
-/* Base journal head number */
-#define BASEHD 1
-/* First "general purpose" journal head */
-#define DATAHD 2
+/* Shorter names for journal head numbers for internal usage */
+#define GCHD UBIFS_GC_HEAD
+#define BASEHD UBIFS_BASE_HEAD
+#define DATAHD UBIFS_DATA_HEAD
/* 'No change' value for 'ubifs_change_lp()' */
#define LPROPS_NC 0x80000001
/* scan.c */
struct ubifs_scan_leb *ubifs_scan(const struct ubifs_info *c, int lnum,
- int offs, void *sbuf);
+ int offs, void *sbuf, int quiet);
void ubifs_scan_destroy(struct ubifs_scan_leb *sleb);
int ubifs_scan_a_node(const struct ubifs_info *c, void *buf, int len, int lnum,
int offs, int quiet);
const struct ubifs_lprops *ubifs_fast_find_empty(struct ubifs_info *c);
const struct ubifs_lprops *ubifs_fast_find_freeable(struct ubifs_info *c);
const struct ubifs_lprops *ubifs_fast_find_frdi_idx(struct ubifs_info *c);
+int ubifs_calc_dark(const struct ubifs_info *c, int spc);
/* file.c */
int ubifs_fsync(struct file *file, struct dentry *dentry, int datasync);
SECURITY_XATTR,
};
-static struct inode_operations none_inode_operations;
+static const struct inode_operations none_inode_operations;
static struct address_space_operations none_address_operations;
-static struct file_operations none_file_operations;
+static const struct file_operations none_file_operations;
/**
* create_xattr - create an extended attribute.
git://ftp.safe.ca / safe / jmp / linux-2.6 / commitdiff