From 171bbfbeab7730031eec8025341401fabe540bd5 Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Tue, 25 Nov 2008 17:42:31 -0500 Subject: jbd2: Add BH_JBDPrivateStart Add this so that file systems using JBD2 can safely allocate unused b_state bits. In this case, we add it so that Ocfs2 can define a single bit for tracking the validation state of a buffer. Signed-off-by: Mark Fasheh Signed-off-by: "Theodore Ts'o" --- include/linux/jbd2.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/jbd2.h') diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index c7d106ef22e..f3664574548 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -329,6 +329,7 @@ enum jbd_state_bits { BH_State, /* Pins most journal_head state */ BH_JournalHead, /* Pins bh->b_private and jh->b_bh */ BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */ + BH_JBDPrivateStart, /* First bit available for private use by FS */ }; BUFFER_FNS(JBD, jbd) -- cgit v1.2.3 From e07f7183a486cf9783d1f8c9d2997b5b39eeb2d4 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 26 Nov 2008 01:14:26 -0500 Subject: jbd2: improve jbd2 fsync batching This patch removes the static sleep time in favor of a more self optimizing approach where we measure the average amount of time it takes to commit a transaction to disk and the ammount of time a transaction has been running. If somebody does a sync write or an fsync() traditionally we would sleep for 1 jiffies, which depending on the value of HZ could be a significant amount of time compared to how long it takes to commit a transaction to the underlying storage. With this patch instead of sleeping for a jiffie, we check to see if the amount of time this transaction has been running is less than the average commit time, and if it is we sleep for the delta using schedule_hrtimeout to give us a higher precision sleep time. This greatly benefits high end storage where you could end up sleeping for longer than it takes to commit the transaction and therefore sitting idle instead of allowing the transaction to be committed by keeping the sleep time to a minimum so you are sure to always be doing something. Signed-off-by: Josef Bacik Signed-off-by: "Theodore Ts'o" --- include/linux/jbd2.h | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'include/linux/jbd2.h') diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index f3664574548..ab8cef130c2 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -637,6 +637,11 @@ struct transaction_s */ unsigned long t_expires; + /* + * When this transaction started, in nanoseconds [no locking] + */ + ktime_t t_start_time; + /* * How many handles used this transaction? [t_handle_lock] */ @@ -939,8 +944,18 @@ struct journal_s struct buffer_head **j_wbuf; int j_wbufsize; + /* + * this is the pid of hte last person to run a synchronous operation + * through the journal + */ pid_t j_last_sync_writer; + /* + * the average amount of time in nanoseconds it takes to commit a + * transaction to disk. [j_state_lock] + */ + u64 j_average_commit_time; + /* This function is called when a transaction is closed */ void (*j_commit_callback)(journal_t *, transaction_t *); -- cgit v1.2.3 From 30773840c19cea60dcef39545960d541b1ac1cf8 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 3 Jan 2009 20:27:38 -0500 Subject: ext4: add fsync batch tuning knobs Add new mount options, min_batch_time and max_batch_time, which controls how long the jbd2 layer should wait for additional filesystem operations to get batched with a synchronous write transaction. Signed-off-by: "Theodore Ts'o" --- include/linux/jbd2.h | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'include/linux/jbd2.h') diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index ab8cef130c2..a3cd647ea1b 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -956,6 +956,14 @@ struct journal_s */ u64 j_average_commit_time; + /* + * minimum and maximum times that we should wait for + * additional filesystem operations to get batched into a + * synchronous handle in microseconds + */ + u32 j_min_batch_time; + u32 j_max_batch_time; + /* This function is called when a transaction is closed */ void (*j_commit_callback)(journal_t *, transaction_t *); -- cgit v1.2.3 From 1a0d3786dd57dbd74f340322054c3d618b999dcf Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 5 Nov 2008 00:09:22 -0500 Subject: jbd2: Remove a large array of bh's from the stack of the checkpoint routine jbd2_log_do_checkpoint()n is one of the kernel's largest stack users. Move the array of buffer head's from the stack of jbd2_log_do_checkpoint() to the in-core journal structure. Signed-off-by: "Theodore Ts'o" --- include/linux/jbd2.h | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'include/linux/jbd2.h') diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index a3cd647ea1b..004c9a8d63e 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -687,6 +687,8 @@ jbd2_time_diff(unsigned long start, unsigned long end) return end + (MAX_JIFFY_OFFSET - start); } +#define JBD2_NR_BATCH 64 + /** * struct journal_s - The journal_s type is the concrete type associated with * journal_t. @@ -830,6 +832,14 @@ struct journal_s /* Semaphore for locking against concurrent checkpoints */ struct mutex j_checkpoint_mutex; + /* + * List of buffer heads used by the checkpoint routine. This + * was moved from jbd2_log_do_checkpoint() to reduce stack + * usage. Access to this array is controlled by the + * j_checkpoint_mutex. [j_checkpoint_mutex] + */ + struct buffer_head *j_chkpt_bhs[JBD2_NR_BATCH]; + /* * Journal head: identifies the first unused block in the journal. * [j_state_lock] -- cgit v1.2.3 From fb68407b0d9efba962c03f55009c797e22f024bc Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Thu, 6 Nov 2008 17:50:21 -0500 Subject: jbd2: Call journal commit callback without holding j_list_lock Avoid freeing the transaction in __jbd2_journal_drop_transaction() so the journal commit callback can run without holding j_list_lock, to avoid lock contention on this spinlock. Signed-off-by: Aneesh Kumar K.V Signed-off-by: "Theodore Ts'o" --- include/linux/jbd2.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux/jbd2.h') diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index 004c9a8d63e..9d82084a160 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -1179,8 +1179,8 @@ int jbd2_log_wait_commit(journal_t *journal, tid_t tid); int jbd2_log_do_checkpoint(journal_t *journal); void __jbd2_log_wait_for_space(journal_t *journal); -extern void __jbd2_journal_drop_transaction(journal_t *, transaction_t *); -extern int jbd2_cleanup_journal_tail(journal_t *); +extern void __jbd2_journal_drop_transaction(journal_t *, transaction_t *); +extern int jbd2_cleanup_journal_tail(journal_t *); /* Debugging code only: */ -- cgit v1.2.3 From c31910672376dfb8d020e32afa7249763bcd924a Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Tue, 6 Jan 2009 11:14:25 -0500 Subject: ext4: Remove code to create the journal inode This code has been obsolete in quite some time, since the supported method for adding a journal inode is to use tune2fs (or to creating new filesystem with a journal via mke2fs or mkfs.ext4). Signed-off-by: "Theodore Ts'o" --- include/linux/jbd2.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/linux/jbd2.h') diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index 9d82084a160..adef1c9940d 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -1104,7 +1104,6 @@ extern int jbd2_journal_set_features (journal_t *, unsigned long, unsigned long, unsigned long); extern void jbd2_journal_clear_features (journal_t *, unsigned long, unsigned long, unsigned long); -extern int jbd2_journal_create (journal_t *); extern int jbd2_journal_load (journal_t *journal); extern int jbd2_journal_destroy (journal_t *); extern int jbd2_journal_recover (journal_t *journal); -- cgit v1.2.3