6 files changed, 92 insertions, 81 deletions

diff --git a/fs/jbd/checkpoint.c b/fs/jbd/checkpoint.c
index 61f32f3868cd..b0435dd0654d 100644
--- a/fs/jbd/checkpoint.c
+++ b/fs/jbd/checkpoint.c
@@ -456,7 +456,7 @@ int cleanup_journal_tail(journal_t *journal)
 {
         transaction_t * transaction;
         tid_t           first_tid;
-        unsigned long   blocknr, freed;
+        unsigned int    blocknr, freed;
 
         if (is_journal_aborted(journal))
                 return 1;
@@ -502,8 +502,8 @@ int cleanup_journal_tail(journal_t *journal)
                 freed = freed + journal->j_last - journal->j_first;
 
         jbd_debug(1,
-                  "Cleaning journal tail from %d to %d (offset %lu), "
-                  "freeing %lu\n",
+                  "Cleaning journal tail from %d to %d (offset %u), "
+                  "freeing %u\n",
                   journal->j_tail_sequence, first_tid, blocknr, freed);
 
         journal->j_free += freed;
diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c
index 618e21c0b7a3..4bd882548c45 100644
--- a/fs/jbd/commit.c
+++ b/fs/jbd/commit.c
@@ -308,7 +308,7 @@ void journal_commit_transaction(journal_t *journal)
         int bufs;
         int flags;
         int err;
-        unsigned long blocknr;
+        unsigned int blocknr;
         ktime_t start_time;
         u64 commit_time;
         char *tagp = NULL;
diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c
index 737f7246a4b5..bd3c073b485d 100644
--- a/fs/jbd/journal.c
+++ b/fs/jbd/journal.c
@@ -276,7 +276,7 @@ static void journal_kill_thread(journal_t *journal)
 int journal_write_metadata_buffer(transaction_t *transaction,
                                   struct journal_head  *jh_in,
                                   struct journal_head **jh_out,
-                                  unsigned long blocknr)
+                                  unsigned int blocknr)
 {
         int need_copy_out = 0;
         int done_copy_out = 0;
@@ -287,6 +287,7 @@ int journal_write_metadata_buffer(transaction_t *transaction,
         struct page *new_page;
         unsigned int new_offset;
         struct buffer_head *bh_in = jh2bh(jh_in);
+        journal_t *journal = transaction->t_journal;
 
         /*
          * The buffer really shouldn't be locked: only the current committing
@@ -300,6 +301,11 @@ int journal_write_metadata_buffer(transaction_t *transaction,
         J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in));
 
         new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL);
+        /* keep subsequent assertions sane */
+        new_bh->b_state = 0;
+        init_buffer(new_bh, NULL, NULL);
+        atomic_set(&new_bh->b_count, 1);
+        new_jh = journal_add_journal_head(new_bh);      /* This sleeps */
 
         /*
          * If a new transaction has already done a buffer copy-out, then
@@ -361,14 +367,6 @@ repeat:
                 kunmap_atomic(mapped_data, KM_USER0);
         }
 
-        /* keep subsequent assertions sane */
-        new_bh->b_state = 0;
-        init_buffer(new_bh, NULL, NULL);
-        atomic_set(&new_bh->b_count, 1);
-        jbd_unlock_bh_state(bh_in);
-
-        new_jh = journal_add_journal_head(new_bh);      /* This sleeps */
-
         set_bh_page(new_bh, new_page, new_offset);
         new_jh->b_transaction = NULL;
         new_bh->b_size = jh2bh(jh_in)->b_size;
@@ -385,7 +383,11 @@ repeat:
          * copying is moved to the transaction's shadow queue.
          */
         JBUFFER_TRACE(jh_in, "file as BJ_Shadow");
-        journal_file_buffer(jh_in, transaction, BJ_Shadow);
+        spin_lock(&journal->j_list_lock);
+        __journal_file_buffer(jh_in, transaction, BJ_Shadow);
+        spin_unlock(&journal->j_list_lock);
+        jbd_unlock_bh_state(bh_in);
+
         JBUFFER_TRACE(new_jh, "file as BJ_IO");
         journal_file_buffer(new_jh, transaction, BJ_IO);
 
@@ -565,9 +567,9 @@ int log_wait_commit(journal_t *journal, tid_t tid)
  * Log buffer allocation routines:
  */
 
-int journal_next_log_block(journal_t *journal, unsigned long *retp)
+int journal_next_log_block(journal_t *journal, unsigned int *retp)
 {
-        unsigned long blocknr;
+        unsigned int blocknr;
 
         spin_lock(&journal->j_state_lock);
         J_ASSERT(journal->j_free > 1);
@@ -588,11 +590,11 @@ int journal_next_log_block(journal_t *journal, unsigned long *retp)
  * this is a no-op.  If needed, we can use j_blk_offset - everything is
  * ready.
  */
-int journal_bmap(journal_t *journal, unsigned long blocknr,
-                 unsigned long *retp)
+int journal_bmap(journal_t *journal, unsigned int blocknr,
+                 unsigned int *retp)
 {
         int err = 0;
-        unsigned long ret;
+        unsigned int ret;
 
         if (journal->j_inode) {
                 ret = bmap(journal->j_inode, blocknr);
@@ -602,7 +604,7 @@ int journal_bmap(journal_t *journal, unsigned long blocknr,
                         char b[BDEVNAME_SIZE];
 
                         printk(KERN_ALERT "%s: journal block not found "
-                                        "at offset %lu on %s\n",
+                                        "at offset %u on %s\n",
                                 __func__,
                                 blocknr,
                                 bdevname(journal->j_dev, b));
@@ -628,7 +630,7 @@ int journal_bmap(journal_t *journal, unsigned long blocknr,
 struct journal_head *journal_get_descriptor_buffer(journal_t *journal)
 {
         struct buffer_head *bh;
-        unsigned long blocknr;
+        unsigned int blocknr;
         int err;
 
         err = journal_next_log_block(journal, &blocknr);
@@ -772,7 +774,7 @@ journal_t * journal_init_inode (struct inode *inode)
         journal_t *journal = journal_init_common();
         int err;
         int n;
-        unsigned long blocknr;
+        unsigned int blocknr;
 
         if (!journal)
                 return NULL;
@@ -844,10 +846,16 @@ static void journal_fail_superblock (journal_t *journal)
 static int journal_reset(journal_t *journal)
 {
         journal_superblock_t *sb = journal->j_superblock;
-        unsigned long first, last;
+        unsigned int first, last;
 
         first = be32_to_cpu(sb->s_first);
         last = be32_to_cpu(sb->s_maxlen);
+        if (first + JFS_MIN_JOURNAL_BLOCKS > last + 1) {
+                printk(KERN_ERR "JBD: Journal too short (blocks %u-%u).\n",
+                       first, last);
+                journal_fail_superblock(journal);
+                return -EINVAL;
+        }
 
         journal->j_first = first;
         journal->j_last = last;
@@ -877,7 +885,7 @@ static int journal_reset(journal_t *journal)
  **/
 int journal_create(journal_t *journal)
 {
-        unsigned long blocknr;
+        unsigned int blocknr;
         struct buffer_head *bh;
         journal_superblock_t *sb;
         int i, err;
@@ -961,14 +969,14 @@ void journal_update_superblock(journal_t *journal, int wait)
         if (sb->s_start == 0 && journal->j_tail_sequence ==
                                 journal->j_transaction_sequence) {
                 jbd_debug(1,"JBD: Skipping superblock update on recovered sb "
-                        "(start %ld, seq %d, errno %d)\n",
+                        "(start %u, seq %d, errno %d)\n",
                         journal->j_tail, journal->j_tail_sequence,
                         journal->j_errno);
                 goto out;
         }
 
         spin_lock(&journal->j_state_lock);
-        jbd_debug(1,"JBD: updating superblock (start %ld, seq %d, errno %d)\n",
+        jbd_debug(1,"JBD: updating superblock (start %u, seq %d, errno %d)\n",
                   journal->j_tail, journal->j_tail_sequence, journal->j_errno);
 
         sb->s_sequence = cpu_to_be32(journal->j_tail_sequence);
@@ -1363,7 +1371,7 @@ int journal_flush(journal_t *journal)
 {
         int err = 0;
         transaction_t *transaction = NULL;
-        unsigned long old_tail;
+        unsigned int old_tail;
 
         spin_lock(&journal->j_state_lock);
 
diff --git a/fs/jbd/recovery.c b/fs/jbd/recovery.c
index db5e982c5ddf..cb1a49ae605e 100644
--- a/fs/jbd/recovery.c
+++ b/fs/jbd/recovery.c
@@ -70,7 +70,7 @@ static int do_readahead(journal_t *journal, unsigned int start)
 {
         int err;
         unsigned int max, nbufs, next;
-        unsigned long blocknr;
+        unsigned int blocknr;
         struct buffer_head *bh;
 
         struct buffer_head * bufs[MAXBUF];
@@ -132,7 +132,7 @@ static int jread(struct buffer_head **bhp, journal_t *journal,
                  unsigned int offset)
 {
         int err;
-        unsigned long blocknr;
+        unsigned int blocknr;
         struct buffer_head *bh;
 
         *bhp = NULL;
@@ -314,7 +314,7 @@ static int do_one_pass(journal_t *journal,
                         struct recovery_info *info, enum passtype pass)
 {
         unsigned int            first_commit_ID, next_commit_ID;
-        unsigned long           next_log_block;
+        unsigned int            next_log_block;
         int                     err, success = 0;
         journal_superblock_t *  sb;
         journal_header_t *      tmp;
@@ -367,14 +367,14 @@ static int do_one_pass(journal_t *journal,
                         if (tid_geq(next_commit_ID, info->end_transaction))
                                 break;
 
-                jbd_debug(2, "Scanning for sequence ID %u at %lu/%lu\n",
+                jbd_debug(2, "Scanning for sequence ID %u at %u/%u\n",
                           next_commit_ID, next_log_block, journal->j_last);
 
                 /* Skip over each chunk of the transaction looking
                  * either the next descriptor block or the final commit
                  * record. */
 
-                jbd_debug(3, "JBD: checking block %ld\n", next_log_block);
+                jbd_debug(3, "JBD: checking block %u\n", next_log_block);
                 err = jread(&bh, journal, next_log_block);
                 if (err)
                         goto failed;
@@ -429,7 +429,7 @@ static int do_one_pass(journal_t *journal,
                         tagp = &bh->b_data[sizeof(journal_header_t)];
                         while ((tagp - bh->b_data +sizeof(journal_block_tag_t))
                                <= journal->j_blocksize) {
-                                unsigned long io_block;
+                                unsigned int io_block;
 
                                 tag = (journal_block_tag_t *) tagp;
                                 flags = be32_to_cpu(tag->t_flags);
@@ -443,10 +443,10 @@ static int do_one_pass(journal_t *journal,
                                         success = err;
                                         printk (KERN_ERR
                                                 "JBD: IO error %d recovering "
-                                                "block %ld in log\n",
+                                                "block %u in log\n",
                                                 err, io_block);
                                 } else {
-                                        unsigned long blocknr;
+                                        unsigned int blocknr;
 
                                         J_ASSERT(obh != NULL);
                                         blocknr = be32_to_cpu(tag->t_blocknr);
@@ -581,7 +581,7 @@ static int scan_revoke_records(journal_t *journal, struct buffer_head *bh,
         max = be32_to_cpu(header->r_count);
 
         while (offset < max) {
-                unsigned long blocknr;
+                unsigned int blocknr;
                 int err;
 
                 blocknr = be32_to_cpu(* ((__be32 *) (bh->b_data+offset)));
diff --git a/fs/jbd/revoke.c b/fs/jbd/revoke.c
index da6cd9bdaabc..ad717328343a 100644
--- a/fs/jbd/revoke.c
+++ b/fs/jbd/revoke.c
@@ -101,7 +101,7 @@ struct jbd_revoke_record_s
 {
         struct list_head  hash;
         tid_t             sequence;     /* Used for recovery only */
-        unsigned long     blocknr;
+        unsigned int      blocknr;
 };
 
 
@@ -126,7 +126,7 @@ static void flush_descriptor(journal_t *, struct journal_head *, int, int);
 /* Utility functions to maintain the revoke table */
 
 /* Borrowed from buffer.c: this is a tried and tested block hash function */
-static inline int hash(journal_t *journal, unsigned long block)
+static inline int hash(journal_t *journal, unsigned int block)
 {
         struct jbd_revoke_table_s *table = journal->j_revoke;
         int hash_shift = table->hash_shift;
@@ -136,7 +136,7 @@ static inline int hash(journal_t *journal, unsigned long block)
                 (block << (hash_shift - 12))) & (table->hash_size - 1);
 }
 
-static int insert_revoke_hash(journal_t *journal, unsigned long blocknr,
+static int insert_revoke_hash(journal_t *journal, unsigned int blocknr,
                               tid_t seq)
 {
         struct list_head *hash_list;
@@ -166,7 +166,7 @@ oom:
 /* Find a revoke record in the journal's hash table. */
 
 static struct jbd_revoke_record_s *find_revoke_record(journal_t *journal,
-                                                      unsigned long blocknr)
+                                                      unsigned int blocknr)
 {
         struct list_head *hash_list;
         struct jbd_revoke_record_s *record;
@@ -332,7 +332,7 @@ void journal_destroy_revoke(journal_t *journal)
  * by one.
  */
 
-int journal_revoke(handle_t *handle, unsigned long blocknr,
+int journal_revoke(handle_t *handle, unsigned int blocknr,
                    struct buffer_head *bh_in)
 {
         struct buffer_head *bh = NULL;
@@ -401,7 +401,7 @@ int journal_revoke(handle_t *handle, unsigned long blocknr,
                 }
         }
 
-        jbd_debug(2, "insert revoke for block %lu, bh_in=%p\n", blocknr, bh_in);
+        jbd_debug(2, "insert revoke for block %u, bh_in=%p\n", blocknr, bh_in);
         err = insert_revoke_hash(journal, blocknr,
                                 handle->h_transaction->t_tid);
         BUFFER_TRACE(bh_in, "exit");
@@ -644,7 +644,7 @@ static void flush_descriptor(journal_t *journal,
  */
 
 int journal_set_revoke(journal_t *journal,
-                       unsigned long blocknr,
+                       unsigned int blocknr,
                        tid_t sequence)
 {
         struct jbd_revoke_record_s *record;
@@ -668,7 +668,7 @@ int journal_set_revoke(journal_t *journal,
  */
 
 int journal_test_revoke(journal_t *journal,
-                        unsigned long blocknr,
+                        unsigned int blocknr,
                         tid_t sequence)
 {
         struct jbd_revoke_record_s *record;
diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c
index 73242ba7c7b1..006f9ad838a2 100644
--- a/fs/jbd/transaction.c
+++ b/fs/jbd/transaction.c
@@ -56,7 +56,8 @@ get_transaction(journal_t *journal, transaction_t *transaction)
         spin_lock_init(&transaction->t_handle_lock);
 
         /* Set up the commit timer for the new transaction. */
-        journal->j_commit_timer.expires = round_jiffies(transaction->t_expires);
+        journal->j_commit_timer.expires =
+                                round_jiffies_up(transaction->t_expires);
         add_timer(&journal->j_commit_timer);
 
         J_ASSERT(journal->j_running_transaction == NULL);
@@ -228,6 +229,8 @@ repeat_locked:
                   __log_space_left(journal));
         spin_unlock(&transaction->t_handle_lock);
         spin_unlock(&journal->j_state_lock);
+
+        lock_map_acquire(&handle->h_lockdep_map);
 out:
         if (unlikely(new_transaction))          /* It's usually NULL */
                 kfree(new_transaction);
@@ -292,9 +295,6 @@ handle_t *journal_start(journal_t *journal, int nblocks)
                 handle = ERR_PTR(err);
                 goto out;
         }
-
-        lock_map_acquire(&handle->h_lockdep_map);
-
 out:
         return handle;
 }
@@ -416,6 +416,7 @@ int journal_restart(handle_t *handle, int nblocks)
         __log_start_commit(journal, transaction->t_tid);
         spin_unlock(&journal->j_state_lock);
 
+        lock_map_release(&handle->h_lockdep_map);
         handle->h_buffer_credits = nblocks;
         ret = start_this_handle(journal, handle);
         return ret;
@@ -489,34 +490,15 @@ void journal_unlock_updates (journal_t *journal)
         wake_up(&journal->j_wait_transaction_locked);
 }
 
-/*
- * Report any unexpected dirty buffers which turn up.  Normally those
- * indicate an error, but they can occur if the user is running (say)
- * tune2fs to modify the live filesystem, so we need the option of
- * continuing as gracefully as possible.  #
- *
- * The caller should already hold the journal lock and
- * j_list_lock spinlock: most callers will need those anyway
- * in order to probe the buffer's journaling state safely.
- */
-static void jbd_unexpected_dirty_buffer(struct journal_head *jh)
+static void warn_dirty_buffer(struct buffer_head *bh)
 {
-        int jlist;
-
-        /* If this buffer is one which might reasonably be dirty
-         * --- ie. data, or not part of this journal --- then
-         * we're OK to leave it alone, but otherwise we need to
-         * move the dirty bit to the journal's own internal
-         * JBDDirty bit. */
-        jlist = jh->b_jlist;
-
-        if (jlist == BJ_Metadata || jlist == BJ_Reserved ||
-            jlist == BJ_Shadow || jlist == BJ_Forget) {
-                struct buffer_head *bh = jh2bh(jh);
+        char b[BDEVNAME_SIZE];
 
-                if (test_clear_buffer_dirty(bh))
-                        set_buffer_jbddirty(bh);
-        }
+        printk(KERN_WARNING
+               "JBD: Spotted dirty metadata buffer (dev = %s, blocknr = %llu). "
+               "There's a risk of filesystem corruption in case of system "
+               "crash.\n",
+               bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr);
 }
 
 /*
@@ -583,14 +565,16 @@ repeat:
                         if (jh->b_next_transaction)
                                 J_ASSERT_JH(jh, jh->b_next_transaction ==
                                                         transaction);
+                        warn_dirty_buffer(bh);
                 }
                 /*
                  * In any case we need to clean the dirty flag and we must
                  * do it under the buffer lock to be sure we don't race
                  * with running write-out.
                  */
-                JBUFFER_TRACE(jh, "Unexpected dirty buffer");
-                jbd_unexpected_dirty_buffer(jh);
+                JBUFFER_TRACE(jh, "Journalling dirty buffer");
+                clear_buffer_dirty(bh);
+                set_buffer_jbddirty(bh);
         }
 
         unlock_buffer(bh);
@@ -826,6 +810,15 @@ int journal_get_create_access(handle_t *handle, struct buffer_head *bh)
         J_ASSERT_JH(jh, buffer_locked(jh2bh(jh)));
 
         if (jh->b_transaction == NULL) {
+                /*
+                 * Previous journal_forget() could have left the buffer
+                 * with jbddirty bit set because it was being committed. When
+                 * the commit finished, we've filed the buffer for
+                 * checkpointing and marked it dirty. Now we are reallocating
+                 * the buffer so the transaction freeing it must have
+                 * committed and so it's safe to clear the dirty bit.
+                 */
+                clear_buffer_dirty(jh2bh(jh));
                 jh->b_transaction = transaction;
 
                 /* first access by this transaction */
@@ -1782,8 +1775,13 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction)
 
         if (jh->b_cp_transaction) {
                 JBUFFER_TRACE(jh, "on running+cp transaction");
+                /*
+                 * We don't want to write the buffer anymore, clear the
+                 * bit so that we don't confuse checks in
+                 * __journal_file_buffer
+                 */
+                clear_buffer_dirty(bh);
                 __journal_file_buffer(jh, transaction, BJ_Forget);
-                clear_buffer_jbddirty(bh);
                 may_free = 0;
         } else {
                 JBUFFER_TRACE(jh, "on running transaction");
@@ -2041,12 +2039,17 @@ void __journal_file_buffer(struct journal_head *jh,
         if (jh->b_transaction && jh->b_jlist == jlist)
                 return;
 
-        /* The following list of buffer states needs to be consistent
-         * with __jbd_unexpected_dirty_buffer()'s handling of dirty
-         * state. */
-
         if (jlist == BJ_Metadata || jlist == BJ_Reserved ||
             jlist == BJ_Shadow || jlist == BJ_Forget) {
+                /*
+                 * For metadata buffers, we track dirty bit in buffer_jbddirty
+                 * instead of buffer_dirty. We should not see a dirty bit set
+                 * here because we clear it in do_get_write_access but e.g.
+                 * tune2fs can modify the sb and set the dirty bit at any time
+                 * so we try to gracefully handle that.
+                 */
+                if (buffer_dirty(bh))
+                        warn_dirty_buffer(bh);
                 if (test_clear_buffer_dirty(bh) ||
                     test_clear_buffer_jbddirty(bh))
                         was_dirty = 1;