Merge commit 'v2.6.38-rc1' into kbuild/packaging

6 files changed, 134 insertions, 89 deletions

diff --git a/fs/jbd/checkpoint.c b/fs/jbd/checkpoint.c
index b0435dd0654d..e4b87bc1fa56 100644
--- a/fs/jbd/checkpoint.c
+++ b/fs/jbd/checkpoint.c
@@ -221,7 +221,7 @@ restart:
                         goto restart;
                 }
                 if (buffer_locked(bh)) {
-                        atomic_inc(&bh->b_count);
+                        get_bh(bh);
                         spin_unlock(&journal->j_list_lock);
                         jbd_unlock_bh_state(bh);
                         wait_on_buffer(bh);
@@ -254,7 +254,9 @@ __flush_batch(journal_t *journal, struct buffer_head **bhs, int *batch_count)
 {
         int i;
 
-        ll_rw_block(SWRITE, *batch_count, bhs);
+        for (i = 0; i < *batch_count; i++)
+                write_dirty_buffer(bhs[i], WRITE);
+
         for (i = 0; i < *batch_count; i++) {
                 struct buffer_head *bh = bhs[i];
                 clear_buffer_jwrite(bh);
@@ -281,7 +283,7 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh,
         int ret = 0;
 
         if (buffer_locked(bh)) {
-                atomic_inc(&bh->b_count);
+                get_bh(bh);
                 spin_unlock(&journal->j_list_lock);
                 jbd_unlock_bh_state(bh);
                 wait_on_buffer(bh);
diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c
index 4bd882548c45..34a4861c14b8 100644
--- a/fs/jbd/commit.c
+++ b/fs/jbd/commit.c
@@ -17,7 +17,6 @@
 #include <linux/fs.h>
 #include <linux/jbd.h>
 #include <linux/errno.h>
-#include <linux/slab.h>
 #include <linux/mm.h>
 #include <linux/pagemap.h>
 #include <linux/bio.h>
@@ -120,7 +119,6 @@ static int journal_write_commit_record(journal_t *journal,
         struct buffer_head *bh;
         journal_header_t *header;
         int ret;
-        int barrier_done = 0;
 
         if (is_journal_aborted(journal))
                 return 0;
@@ -138,34 +136,12 @@ static int journal_write_commit_record(journal_t *journal,
 
         JBUFFER_TRACE(descriptor, "write commit block");
         set_buffer_dirty(bh);
-        if (journal->j_flags & JFS_BARRIER) {
-                set_buffer_ordered(bh);
-                barrier_done = 1;
-        }
-        ret = sync_dirty_buffer(bh);
-        if (barrier_done)
-                clear_buffer_ordered(bh);
-        /* is it possible for another commit to fail at roughly
-         * the same time as this one?  If so, we don't want to
-         * trust the barrier flag in the super, but instead want
-         * to remember if we sent a barrier request
-         */
-        if (ret == -EOPNOTSUPP && barrier_done) {
-                char b[BDEVNAME_SIZE];
-
-                printk(KERN_WARNING
-                        "JBD: barrier-based sync failed on %s - "
-                        "disabling barriers\n",
-                        bdevname(journal->j_dev, b));
-                spin_lock(&journal->j_state_lock);
-                journal->j_flags &= ~JFS_BARRIER;
-                spin_unlock(&journal->j_state_lock);
 
-                /* And try again, without the barrier */
-                set_buffer_uptodate(bh);
-                set_buffer_dirty(bh);
+        if (journal->j_flags & JFS_BARRIER)
+                ret = __sync_dirty_buffer(bh, WRITE_SYNC | WRITE_FLUSH_FUA);
+        else
                 ret = sync_dirty_buffer(bh);
-        }
+
         put_bh(bh);             /* One for getblk() */
         journal_put_journal_head(descriptor);
 
@@ -318,7 +294,7 @@ void journal_commit_transaction(journal_t *journal)
         int first_tag = 0;
         int tag_flag;
         int i;
-        int write_op = WRITE;
+        int write_op = WRITE_SYNC;
 
         /*
          * First job: lock down the current transaction and wait for
@@ -611,13 +587,13 @@ void journal_commit_transaction(journal_t *journal)
                 /* Bump b_count to prevent truncate from stumbling over
                    the shadowed buffer!  @@@ This can go if we ever get
                    rid of the BJ_IO/BJ_Shadow pairing of buffers. */
-                atomic_inc(&jh2bh(jh)->b_count);
+                get_bh(jh2bh(jh));
 
                 /* Make a temporary IO buffer with which to write it out
                    (this will requeue both the metadata buffer and the
                    temporary IO buffer). new_bh goes on BJ_IO*/
 
-                set_bit(BH_JWrite, &jh2bh(jh)->b_state);
+                set_buffer_jwrite(jh2bh(jh));
                 /*
                  * akpm: journal_write_metadata_buffer() sets
                  * new_bh->b_transaction to commit_transaction.
@@ -627,7 +603,7 @@ void journal_commit_transaction(journal_t *journal)
                 JBUFFER_TRACE(jh, "ph3: write metadata");
                 flags = journal_write_metadata_buffer(commit_transaction,
                                                       jh, &new_jh, blocknr);
-                set_bit(BH_JWrite, &jh2bh(new_jh)->b_state);
+                set_buffer_jwrite(jh2bh(new_jh));
                 wbuf[bufs++] = jh2bh(new_jh);
 
                 /* Record the new block's tag in the current descriptor
@@ -737,7 +713,7 @@ wait_for_iobuf:
                    shadowed buffer */
                 jh = commit_transaction->t_shadow_list->b_tprev;
                 bh = jh2bh(jh);
-                clear_bit(BH_JWrite, &bh->b_state);
+                clear_buffer_jwrite(bh);
                 J_ASSERT_BH(bh, buffer_jbddirty(bh));
 
                 /* The metadata is now released for reuse, but we need
@@ -787,6 +763,12 @@ wait_for_iobuf:
 
         jbd_debug(3, "JBD: commit phase 6\n");
 
+        /* All metadata is written, now write commit record and do cleanup */
+        spin_lock(&journal->j_state_lock);
+        J_ASSERT(commit_transaction->t_state == T_COMMIT);
+        commit_transaction->t_state = T_COMMIT_RECORD;
+        spin_unlock(&journal->j_state_lock);
+
         if (journal_write_commit_record(journal, commit_transaction))
                 err = -EIO;
 
@@ -862,12 +844,12 @@ restart_loop:
                 /* A buffer which has been freed while still being
                  * journaled by a previous transaction may end up still
                  * being dirty here, but we want to avoid writing back
-                 * that buffer in the future now that the last use has
-                 * been committed.  That's not only a performance gain,
-                 * it also stops aliasing problems if the buffer is left
-                 * behind for writeback and gets reallocated for another
+                 * that buffer in the future after the "add to orphan"
+                 * operation been committed,  That's not only a performance
+                 * gain, it also stops aliasing problems if the buffer is
+                 * left behind for writeback and gets reallocated for another
                  * use in a different page. */
-                if (buffer_freed(bh)) {
+                if (buffer_freed(bh) && !jh->b_next_transaction) {
                         clear_buffer_freed(bh);
                         clear_buffer_jbddirty(bh);
                 }
@@ -924,7 +906,7 @@ restart_loop:
 
         jbd_debug(3, "JBD: commit phase 8\n");
 
-        J_ASSERT(commit_transaction->t_state == T_COMMIT);
+        J_ASSERT(commit_transaction->t_state == T_COMMIT_RECORD);
 
         commit_transaction->t_state = T_FINISHED;
         J_ASSERT(commit_transaction == journal->j_committing_transaction);
diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c
index bd224eec9b07..da1b5e4ffce1 100644
--- a/fs/jbd/journal.c
+++ b/fs/jbd/journal.c
@@ -36,6 +36,7 @@
 #include <linux/poison.h>
 #include <linux/proc_fs.h>
 #include <linux/debugfs.h>
+#include <linux/ratelimit.h>
 
 #include <asm/uaccess.h>
 #include <asm/page.h>
@@ -84,6 +85,7 @@ EXPORT_SYMBOL(journal_force_commit);
 
 static int journal_convert_superblock_v1(journal_t *, journal_superblock_t *);
 static void __journal_abort_soft (journal_t *journal, int errno);
+static const char *journal_dev_name(journal_t *journal, char *buffer);
 
 /*
  * Helper function used to manage commit timeouts
@@ -439,7 +441,7 @@ int __log_start_commit(journal_t *journal, tid_t target)
          */
         if (!tid_geq(journal->j_commit_request, target)) {
                 /*
-                 * We want a new commit: OK, mark the request and wakup the
+                 * We want a new commit: OK, mark the request and wakeup the
                  * commit thread.  We do _not_ do the commit ourselves.
                  */
 
@@ -565,6 +567,38 @@ int log_wait_commit(journal_t *journal, tid_t tid)
 }
 
 /*
+ * Return 1 if a given transaction has not yet sent barrier request
+ * connected with a transaction commit. If 0 is returned, transaction
+ * may or may not have sent the barrier. Used to avoid sending barrier
+ * twice in common cases.
+ */
+int journal_trans_will_send_data_barrier(journal_t *journal, tid_t tid)
+{
+        int ret = 0;
+        transaction_t *commit_trans;
+
+        if (!(journal->j_flags & JFS_BARRIER))
+                return 0;
+        spin_lock(&journal->j_state_lock);
+        /* Transaction already committed? */
+        if (tid_geq(journal->j_commit_sequence, tid))
+                goto out;
+        /*
+         * Transaction is being committed and we already proceeded to
+         * writing commit record?
+         */
+        commit_trans = journal->j_committing_transaction;
+        if (commit_trans && commit_trans->t_tid == tid &&
+            commit_trans->t_state >= T_COMMIT_RECORD)
+                goto out;
+        ret = 1;
+out:
+        spin_unlock(&journal->j_state_lock);
+        return ret;
+}
+EXPORT_SYMBOL(journal_trans_will_send_data_barrier);
+
+/*
  * Log buffer allocation routines:
  */
 
@@ -918,6 +952,8 @@ int journal_create(journal_t *journal)
                 if (err)
                         return err;
                 bh = __getblk(journal->j_dev, blocknr, journal->j_blocksize);
+                if (unlikely(!bh))
+                        return -ENOMEM;
                 lock_buffer(bh);
                 memset (bh->b_data, 0, journal->j_blocksize);
                 BUFFER_TRACE(bh, "marking dirty");
@@ -978,6 +1014,23 @@ void journal_update_superblock(journal_t *journal, int wait)
                 goto out;
         }
 
+        if (buffer_write_io_error(bh)) {
+                char b[BDEVNAME_SIZE];
+                /*
+                 * Oh, dear.  A previous attempt to write the journal
+                 * superblock failed.  This could happen because the
+                 * USB device was yanked out.  Or it could happen to
+                 * be a transient write error and maybe the block will
+                 * be remapped.  Nothing we can do but to retry the
+                 * write and hope for the best.
+                 */
+                printk(KERN_ERR "JBD: previous I/O error detected "
+                       "for journal superblock update for %s.\n",
+                       journal_dev_name(journal, b));
+                clear_buffer_write_io_error(bh);
+                set_buffer_uptodate(bh);
+        }
+
         spin_lock(&journal->j_state_lock);
         jbd_debug(1,"JBD: updating superblock (start %u, seq %d, errno %d)\n",
                   journal->j_tail, journal->j_tail_sequence, journal->j_errno);
@@ -989,10 +1042,18 @@ void journal_update_superblock(journal_t *journal, int wait)
 
         BUFFER_TRACE(bh, "marking dirty");
         mark_buffer_dirty(bh);
-        if (wait)
+        if (wait) {
                 sync_dirty_buffer(bh);
-        else
-                ll_rw_block(SWRITE, 1, &bh);
+                if (buffer_write_io_error(bh)) {
+                        char b[BDEVNAME_SIZE];
+                        printk(KERN_ERR "JBD: I/O error detected "
+                               "when updating journal superblock for %s.\n",
+                               journal_dev_name(journal, b));
+                        clear_buffer_write_io_error(bh);
+                        set_buffer_uptodate(bh);
+                }
+        } else
+                write_dirty_buffer(bh, WRITE);
 
 out:
         /* If we have just flushed the log (by marking s_start==0), then
@@ -1157,6 +1218,7 @@ int journal_destroy(journal_t *journal)
 {
         int err = 0;
 
+        
         /* Wait for the commit thread to wake up and die. */
         journal_kill_thread(journal);
 
@@ -1248,13 +1310,9 @@ int journal_check_used_features (journal_t *journal, unsigned long compat,
 int journal_check_available_features (journal_t *journal, unsigned long compat,
                                       unsigned long ro, unsigned long incompat)
 {
-        journal_superblock_t *sb;
-
         if (!compat && !ro && !incompat)
                 return 1;
 
-        sb = journal->j_superblock;
-
         /* We can support any known requested features iff the
          * superblock is in version 2.  Otherwise we fail to support any
          * extended sb features. */
@@ -1448,7 +1506,6 @@ int journal_flush(journal_t *journal)
 
 int journal_wipe(journal_t *journal, int write)
 {
-        journal_superblock_t *sb;
         int err = 0;
 
         J_ASSERT (!(journal->j_flags & JFS_LOADED));
@@ -1457,8 +1514,6 @@ int journal_wipe(journal_t *journal, int write)
         if (err)
                 return err;
 
-        sb = journal->j_superblock;
-
         if (!journal->j_tail)
                 goto no_recovery;
 
@@ -1693,7 +1748,6 @@ static void journal_destroy_journal_head_cache(void)
 static struct journal_head *journal_alloc_journal_head(void)
 {
         struct journal_head *ret;
-        static unsigned long last_warning;
 
 #ifdef CONFIG_JBD_DEBUG
         atomic_inc(&nr_journal_heads);
@@ -1701,11 +1755,9 @@ static struct journal_head *journal_alloc_journal_head(void)
         ret = kmem_cache_alloc(journal_head_cache, GFP_NOFS);
         if (ret == NULL) {
                 jbd_debug(1, "out of memory for journal_head\n");
-                if (time_after(jiffies, last_warning + 5*HZ)) {
-                        printk(KERN_NOTICE "ENOMEM in %s, retrying.\n",
-                               __func__);
-                        last_warning = jiffies;
-                }
+                printk_ratelimited(KERN_NOTICE "ENOMEM in %s, retrying.\n",
+                                   __func__);
+
                 while (ret == NULL) {
                         yield();
                         ret = kmem_cache_alloc(journal_head_cache, GFP_NOFS);
diff --git a/fs/jbd/recovery.c b/fs/jbd/recovery.c
index cb1a49ae605e..5b43e96788e6 100644
--- a/fs/jbd/recovery.c
+++ b/fs/jbd/recovery.c
@@ -20,7 +20,6 @@
 #include <linux/fs.h>
 #include <linux/jbd.h>
 #include <linux/errno.h>
-#include <linux/slab.h>
 #endif
 
 /*
@@ -284,12 +283,9 @@ int journal_recover(journal_t *journal)
 int journal_skip_recovery(journal_t *journal)
 {
         int                     err;
-        journal_superblock_t *  sb;
-
         struct recovery_info    info;
 
         memset (&info, 0, sizeof(info));
-        sb = journal->j_superblock;
 
         err = do_one_pass(journal, &info, PASS_SCAN);
 
@@ -298,11 +294,12 @@ int journal_skip_recovery(journal_t *journal)
                 ++journal->j_transaction_sequence;
         } else {
 #ifdef CONFIG_JBD_DEBUG
-                int dropped = info.end_transaction - be32_to_cpu(sb->s_sequence);
-#endif
+                int dropped = info.end_transaction -
+                              be32_to_cpu(journal->j_superblock->s_sequence);
                 jbd_debug(1,
                           "JBD: ignoring %d transaction%s from the journal.\n",
                           dropped, (dropped == 1) ? "" : "s");
+#endif
                 journal->j_transaction_sequence = ++info.end_transaction;
         }
 
@@ -322,11 +319,6 @@ static int do_one_pass(journal_t *journal,
         unsigned int            sequence;
         int                     blocktype;
 
-        /* Precompute the maximum metadata descriptors in a descriptor block */
-        int                     MAX_BLOCKS_PER_DESC;
-        MAX_BLOCKS_PER_DESC = ((journal->j_blocksize-sizeof(journal_header_t))
-                               / sizeof(journal_block_tag_t));
-
         /*
          * First thing is to establish what we expect to find in the log
          * (in terms of transaction IDs), and where (in terms of log
diff --git a/fs/jbd/revoke.c b/fs/jbd/revoke.c
index ad717328343a..d29018307e2e 100644
--- a/fs/jbd/revoke.c
+++ b/fs/jbd/revoke.c
@@ -617,7 +617,7 @@ static void flush_descriptor(journal_t *journal,
         set_buffer_jwrite(bh);
         BUFFER_TRACE(bh, "write");
         set_buffer_dirty(bh);
-        ll_rw_block((write_op == WRITE) ? SWRITE : SWRITE_SYNC_PLUG, 1, &bh);
+        write_dirty_buffer(bh, write_op);
 }
 #endif
 
diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c
index 006f9ad838a2..5b2e4c30a2a1 100644
--- a/fs/jbd/transaction.c
+++ b/fs/jbd/transaction.c
@@ -207,7 +207,7 @@ repeat_locked:
          * the committing transaction.  Really, we only need to give it
          * committing_transaction->t_outstanding_credits plus "enough" for
          * the log control blocks.
-         * Also, this test is inconsitent with the matching one in
+         * Also, this test is inconsistent with the matching one in
          * journal_extend().
          */
         if (__log_space_left(journal) < jbd_space_needed(journal)) {
@@ -293,9 +293,7 @@ handle_t *journal_start(journal_t *journal, int nblocks)
                 jbd_free_handle(handle);
                 current->journal_info = NULL;
                 handle = ERR_PTR(err);
-                goto out;
         }
-out:
         return handle;
 }
 
@@ -528,7 +526,7 @@ do_get_write_access(handle_t *handle, struct journal_head *jh,
         transaction = handle->h_transaction;
         journal = transaction->t_journal;
 
-        jbd_debug(5, "buffer_head %p, force_copy %d\n", jh, force_copy);
+        jbd_debug(5, "journal_head %p, force_copy %d\n", jh, force_copy);
 
         JBUFFER_TRACE(jh, "entry");
 repeat:
@@ -713,7 +711,7 @@ done:
                 J_EXPECT_JH(jh, buffer_uptodate(jh2bh(jh)),
                             "Possible IO failure.\n");
                 page = jh2bh(jh)->b_page;
-                offset = ((unsigned long) jh2bh(jh)->b_data) & ~PAGE_MASK;
+                offset = offset_in_page(jh2bh(jh)->b_data);
                 source = kmap_atomic(page, KM_USER0);
                 memcpy(jh->b_frozen_data, source+offset, jh2bh(jh)->b_size);
                 kunmap_atomic(source, KM_USER0);
@@ -1398,7 +1396,7 @@ int journal_stop(handle_t *handle)
          * the case where our storage is so fast that it is more optimal to go
          * ahead and force a flush and wait for the transaction to be committed
          * than it is to wait for an arbitrary amount of time for new writers to
-         * join the transaction.  We acheive this by measuring how long it takes
+         * join the transaction.  We achieve this by measuring how long it takes
          * to commit a transaction, and compare it with how long this
          * transaction has been running, and if run time < commit time then we
          * sleep for the delta and commit.  This greatly helps super fast disks
@@ -1864,6 +1862,21 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
         if (!jh)
                 goto zap_buffer_no_jh;
 
+        /*
+         * We cannot remove the buffer from checkpoint lists until the
+         * transaction adding inode to orphan list (let's call it T)
+         * is committed.  Otherwise if the transaction changing the
+         * buffer would be cleaned from the journal before T is
+         * committed, a crash will cause that the correct contents of
+         * the buffer will be lost.  On the other hand we have to
+         * clear the buffer dirty bit at latest at the moment when the
+         * transaction marking the buffer as freed in the filesystem
+         * structures is committed because from that moment on the
+         * buffer can be reallocated and used by a different page.
+         * Since the block hasn't been freed yet but the inode has
+         * already been added to orphan list, it is safe for us to add
+         * the buffer to BJ_Forget list of the newest transaction.
+         */
         transaction = jh->b_transaction;
         if (transaction == NULL) {
                 /* First case: not on any transaction.  If it
@@ -1929,16 +1942,15 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
                         goto zap_buffer;
                 }
                 /*
-                 * If it is committing, we simply cannot touch it.  We
-                 * can remove it's next_transaction pointer from the
-                 * running transaction if that is set, but nothing
-                 * else. */
+                 * The buffer is committing, we simply cannot touch
+                 * it. So we just set j_next_transaction to the
+                 * running transaction (if there is one) and mark
+                 * buffer as freed so that commit code knows it should
+                 * clear dirty bits when it is done with the buffer.
+                 */
                 set_buffer_freed(bh);
-                if (jh->b_next_transaction) {
-                        J_ASSERT(jh->b_next_transaction ==
-                                        journal->j_running_transaction);
-                        jh->b_next_transaction = NULL;
-                }
+                if (journal->j_running_transaction && buffer_jbddirty(bh))
+                        jh->b_next_transaction = journal->j_running_transaction;
                 journal_put_journal_head(jh);
                 spin_unlock(&journal->j_list_lock);
                 jbd_unlock_bh_state(bh);
@@ -2120,7 +2132,7 @@ void journal_file_buffer(struct journal_head *jh,
  */
 void __journal_refile_buffer(struct journal_head *jh)
 {
-        int was_dirty;
+        int was_dirty, jlist;
         struct buffer_head *bh = jh2bh(jh);
 
         J_ASSERT_JH(jh, jbd_is_locked_bh_state(bh));
@@ -2142,8 +2154,13 @@ void __journal_refile_buffer(struct journal_head *jh)
         __journal_temp_unlink_buffer(jh);
         jh->b_transaction = jh->b_next_transaction;
         jh->b_next_transaction = NULL;
-        __journal_file_buffer(jh, jh->b_transaction,
-                                jh->b_modified ? BJ_Metadata : BJ_Reserved);
+        if (buffer_freed(bh))
+                jlist = BJ_Forget;
+        else if (jh->b_modified)
+                jlist = BJ_Metadata;
+        else
+                jlist = BJ_Reserved;
+        __journal_file_buffer(jh, jh->b_transaction, jlist);
         J_ASSERT_JH(jh, jh->b_transaction->t_state == T_RUNNING);
 
         if (was_dirty)