OSDN Git Service

jbd2: Simplify journal_unmap_buffer()
authorThomas Gleixner <tglx@linutronix.de>
Fri, 9 Aug 2019 12:42:27 +0000 (14:42 +0200)
committerTheodore Ts'o <tytso@mit.edu>
Mon, 21 Oct 2019 13:16:45 +0000 (09:16 -0400)
journal_unmap_buffer() checks first whether the buffer head is a journal.
If so it takes locks and then invokes jbd2_journal_grab_journal_head()
followed by another check whether this is journal head buffer.

The double checking is pointless.

Replace the initial check with jbd2_journal_grab_journal_head() which
alredy checks whether the buffer head is actually a journal.

Allows also early access to the journal head pointer for the upcoming
conversion of state lock to a regular spinlock.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: linux-ext4@vger.kernel.org
Cc: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20190809124233.13277-2-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
fs/jbd2/transaction.c

index bee8498..2273b10 100644 (file)
@@ -2199,7 +2199,8 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh,
         * holding the page lock. --sct
         */
 
-       if (!buffer_jbd(bh))
+       jh = jbd2_journal_grab_journal_head(bh);
+       if (!jh)
                goto zap_buffer_unlocked;
 
        /* OK, we have data buffer in journaled mode */
@@ -2207,10 +2208,6 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh,
        jbd_lock_bh_state(bh);
        spin_lock(&journal->j_list_lock);
 
-       jh = jbd2_journal_grab_journal_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)
@@ -2332,7 +2329,6 @@ zap_buffer:
         */
        jh->b_modified = 0;
        jbd2_journal_put_journal_head(jh);
-zap_buffer_no_jh:
        spin_unlock(&journal->j_list_lock);
        jbd_unlock_bh_state(bh);
        write_unlock(&journal->j_state_lock);