[PATCH] jbd: journal_dirty_data re-check for unmapped buffers

When running several fsx's and other filesystem stress tests, we found
cases where an unmapped buffer was still being sent to submit_bh by the
ext3 dirty data journaling code.

I saw this happen in two ways, both related to another thread doing a
truncate which would unmap the buffer in question.

Either we would get into journal_dirty_data with a bh which was already
unmapped (although journal_dirty_data_fn had checked for this earlier, the
state was not locked at that point), or it would get unmapped in the middle
of journal_dirty_data when we dropped locks to call sync_dirty_buffer.

By re-checking for mapped state after we've acquired the bh state lock, we
should avoid these races.  If we find a buffer which is no longer mapped,
we essentially ignore it, because journal_unmap_buffer has already decided
that this buffer can go away.

I've also added tracepoints in these two cases, and made a couple other
tracepoint changes that I found useful in debugging this.

Signed-off-by: Eric Sandeen <esandeen@redhat.com>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c
index d5c63047a8b3f89584963a31c85e22f08ee831c6..4f82bcd63e488a6a1b0ff1b8bb20bb7e2a968b3c 100644 (file)
--- a/fs/jbd/transaction.c
+++ b/fs/jbd/transaction.c
@@ -967,6 +967,13 @@ int journal_dirty_data(handle_t *handle, struct buffer_head *bh)
         */
        jbd_lock_bh_state(bh);
        spin_lock(&journal->j_list_lock);
+
+       /* Now that we have bh_state locked, are we really still mapped? */
+       if (!buffer_mapped(bh)) {
+               JBUFFER_TRACE(jh, "unmapped buffer, bailing out");
+               goto no_journal;
+       }
+
        if (jh->b_transaction) {
                JBUFFER_TRACE(jh, "has transaction");
                if (jh->b_transaction != handle->h_transaction) {
@@ -1028,6 +1035,11 @@ int journal_dirty_data(handle_t *handle, struct buffer_head *bh)
                                sync_dirty_buffer(bh);
                                jbd_lock_bh_state(bh);
                                spin_lock(&journal->j_list_lock);
+                               /* Since we dropped the lock... */
+                               if (!buffer_mapped(bh)) {
+                                       JBUFFER_TRACE(jh, "buffer got unmapped");
+                                       goto no_journal;
+                               }
                                /* The buffer may become locked again at any
                                   time if it is redirtied */
                        }
@@ -1824,6 +1836,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
                        }
                }
        } else if (transaction == journal->j_committing_transaction) {
+               JBUFFER_TRACE(jh, "on committing transaction");
                if (jh->b_jlist == BJ_Locked) {
                        /*
                         * The buffer is on the committing transaction's locked
@@ -1838,7 +1851,6 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
                 * can remove it's next_transaction pointer from the
                 * running transaction if that is set, but nothing
                 * else. */
-               JBUFFER_TRACE(jh, "on committing transaction");
                set_buffer_freed(bh);
                if (jh->b_next_transaction) {
                        J_ASSERT(jh->b_next_transaction ==
@@ -1858,6 +1870,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh)
                 * i_size already for this truncate so recovery will not
                 * expose the disk blocks we are discarding here.) */
                J_ASSERT_JH(jh, transaction == journal->j_running_transaction);
+               JBUFFER_TRACE(jh, "on running transaction");
                may_free = __dispose_buffer(jh, transaction);
        }