[PATCH] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[PATCH] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Theodore Ts'o]

If there are pending blocks to be released after a commit, retrying
the allocation after a journal commit has no hope of helping.  So
track how many pending deleted blocks there might be, and don't retry
if there are no pending blocks.

Reported-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/balloc.c    |  9 ++++++++-
 fs/ext4/ext4.h      |  1 +
 fs/ext4/ext4_jbd2.h | 10 +++++++++-
 fs/ext4/fsync.c     |  5 +++--
 fs/ext4/mballoc.c   | 12 ++++++++++--
 5 files changed, 31 insertions(+), 6 deletions(-)

diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 3020fd7..371ac63 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -603,7 +603,14 @@ int ext4_claim_free_clusters(struct ext4_sb_info *sbi,
  */
 int ext4_should_retry_alloc(struct super_block *sb, int *retries)
 {
-	if (!ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
+	unsigned int pending_blocks;
+
+	spin_lock(&EXT4_SB(sb)->s_md_lock);
+	pending_blocks = EXT4_SB(sb)->s_mb_free_pending;
+	spin_unlock(&EXT4_SB(sb)->s_md_lock);
+
+	if (pending_blocks == 0 ||
+	    !ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
 	    (*retries)++ > 3 ||
 	    !EXT4_SB(sb)->s_journal)
 		return 0;
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index b84aa1c..96c73e6 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1430,6 +1430,7 @@ struct ext4_sb_info {
 	unsigned short *s_mb_offsets;
 	unsigned int *s_mb_maxs;
 	unsigned int s_group_info_size;
+	unsigned int s_mb_free_pending;
 
 	/* tunables */
 	unsigned long s_stripe;
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 09c1ef3..b1d52c1 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -175,6 +175,13 @@ struct ext4_journal_cb_entry {
  * There is no guaranteed calling order of multiple registered callbacks on
  * the same transaction.
  */
+static inline void _ext4_journal_callback_add(handle_t *handle,
+			struct ext4_journal_cb_entry *jce)
+{
+	/* Add the jce to transaction's private list */
+	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
+}
+
 static inline void ext4_journal_callback_add(handle_t *handle,
 			void (*func)(struct super_block *sb,
 				     struct ext4_journal_cb_entry *jce,
@@ -187,10 +194,11 @@ static inline void ext4_journal_callback_add(handle_t *handle,
 	/* Add the jce to transaction's private list */
 	jce->jce_func = func;
 	spin_lock(&sbi->s_md_lock);
-	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
+	_ext4_journal_callback_add(handle, jce);
 	spin_unlock(&sbi->s_md_lock);
 }
 
+
 /**
  * ext4_journal_callback_del: delete a registered callback
  * @handle: active journal transaction handle on which callback was registered
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index 8850254..7ee4abd 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -106,10 +106,10 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	}
 
 	if (!journal) {
-		ret = generic_file_fsync(file, start, end, datasync);
+		ret = __generic_file_fsync(file, start, end, datasync);
 		if (!ret && !hlist_empty(&inode->i_dentry))
 			ret = ext4_sync_parent(inode);
-		goto out;
+		goto issue_flush;
 	}
 
 	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
@@ -140,6 +140,7 @@ int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 		needs_barrier = true;
 	ret = jbd2_complete_transaction(journal, commit_tid);
 	if (needs_barrier) {
+	issue_flush:
 		err = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
 		if (!ret)
 			ret = err;
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index c1ab3ec..77249e1 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2627,6 +2627,7 @@ int ext4_mb_init(struct super_block *sb)
 
 	spin_lock_init(&sbi->s_md_lock);
 	spin_lock_init(&sbi->s_bal_lock);
+	sbi->s_mb_free_pending = 0;
 
 	sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN;
 	sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
@@ -2814,6 +2815,9 @@ static void ext4_free_data_callback(struct super_block *sb,
 	/* we expect to find existing buddy because it's pinned */
 	BUG_ON(err != 0);
 
+	spin_lock(&EXT4_SB(sb)->s_md_lock);
+	EXT4_SB(sb)->s_mb_free_pending -= entry->efd_count;
+	spin_unlock(&EXT4_SB(sb)->s_md_lock);
 
 	db = e4b.bd_info;
 	/* there are blocks to put in buddy to make them really free */
@@ -4583,6 +4587,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 {
 	ext4_group_t group = e4b->bd_group;
 	ext4_grpblk_t cluster;
+	ext4_grpblk_t clusters = new_entry->efd_count;
 	struct ext4_free_data *entry;
 	struct ext4_group_info *db = e4b->bd_info;
 	struct super_block *sb = e4b->bd_sb;
@@ -4649,8 +4654,11 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 		}
 	}
 	/* Add the extent to transaction's private list */
-	ext4_journal_callback_add(handle, ext4_free_data_callback,
-				  &new_entry->efd_jce);
+	new_entry->efd_jce.jce_func = ext4_free_data_callback;
+	spin_lock(&sbi->s_md_lock);
+	_ext4_journal_callback_add(handle, &new_entry->efd_jce);
+	sbi->s_mb_free_pending += clusters;
+	spin_unlock(&sbi->s_md_lock);
 	return 0;
 }
 
-- 
2.5.0

[PATCH -v2] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Theodore Ts'o]

If there are pending blocks to be released after a commit, retrying
the allocation after a journal commit has no hope of helping.  So
track how many pending deleted blocks there might be, and don't retry
if there are no pending blocks.

Reported-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---

Oops, ignore the earlier version of this patch.  I bobbled the commit
and merged in part of another change.

 fs/ext4/balloc.c    |  9 ++++++++-
 fs/ext4/ext4.h      |  1 +
 fs/ext4/ext4_jbd2.h | 10 +++++++++-
 fs/ext4/mballoc.c   | 12 ++++++++++--
 4 files changed, 28 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 3020fd7..371ac63 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -603,7 +603,14 @@ int ext4_claim_free_clusters(struct ext4_sb_info *sbi,
  */
 int ext4_should_retry_alloc(struct super_block *sb, int *retries)
 {
-	if (!ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
+	unsigned int pending_blocks;
+
+	spin_lock(&EXT4_SB(sb)->s_md_lock);
+	pending_blocks = EXT4_SB(sb)->s_mb_free_pending;
+	spin_unlock(&EXT4_SB(sb)->s_md_lock);
+
+	if (pending_blocks == 0 ||
+	    !ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
 	    (*retries)++ > 3 ||
 	    !EXT4_SB(sb)->s_journal)
 		return 0;
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index b84aa1c..96c73e6 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1430,6 +1430,7 @@ struct ext4_sb_info {
 	unsigned short *s_mb_offsets;
 	unsigned int *s_mb_maxs;
 	unsigned int s_group_info_size;
+	unsigned int s_mb_free_pending;
 
 	/* tunables */
 	unsigned long s_stripe;
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 09c1ef3..b1d52c1 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -175,6 +175,13 @@ struct ext4_journal_cb_entry {
  * There is no guaranteed calling order of multiple registered callbacks on
  * the same transaction.
  */
+static inline void _ext4_journal_callback_add(handle_t *handle,
+			struct ext4_journal_cb_entry *jce)
+{
+	/* Add the jce to transaction's private list */
+	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
+}
+
 static inline void ext4_journal_callback_add(handle_t *handle,
 			void (*func)(struct super_block *sb,
 				     struct ext4_journal_cb_entry *jce,
@@ -187,10 +194,11 @@ static inline void ext4_journal_callback_add(handle_t *handle,
 	/* Add the jce to transaction's private list */
 	jce->jce_func = func;
 	spin_lock(&sbi->s_md_lock);
-	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
+	_ext4_journal_callback_add(handle, jce);
 	spin_unlock(&sbi->s_md_lock);
 }
 
+
 /**
  * ext4_journal_callback_del: delete a registered callback
  * @handle: active journal transaction handle on which callback was registered
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index c1ab3ec..77249e1 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2627,6 +2627,7 @@ int ext4_mb_init(struct super_block *sb)
 
 	spin_lock_init(&sbi->s_md_lock);
 	spin_lock_init(&sbi->s_bal_lock);
+	sbi->s_mb_free_pending = 0;
 
 	sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN;
 	sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
@@ -2814,6 +2815,9 @@ static void ext4_free_data_callback(struct super_block *sb,
 	/* we expect to find existing buddy because it's pinned */
 	BUG_ON(err != 0);
 
+	spin_lock(&EXT4_SB(sb)->s_md_lock);
+	EXT4_SB(sb)->s_mb_free_pending -= entry->efd_count;
+	spin_unlock(&EXT4_SB(sb)->s_md_lock);
 
 	db = e4b.bd_info;
 	/* there are blocks to put in buddy to make them really free */
@@ -4583,6 +4587,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 {
 	ext4_group_t group = e4b->bd_group;
 	ext4_grpblk_t cluster;
+	ext4_grpblk_t clusters = new_entry->efd_count;
 	struct ext4_free_data *entry;
 	struct ext4_group_info *db = e4b->bd_info;
 	struct super_block *sb = e4b->bd_sb;
@@ -4649,8 +4654,11 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 		}
 	}
 	/* Add the extent to transaction's private list */
-	ext4_journal_callback_add(handle, ext4_free_data_callback,
-				  &new_entry->efd_jce);
+	new_entry->efd_jce.jce_func = ext4_free_data_callback;
+	spin_lock(&sbi->s_md_lock);
+	_ext4_journal_callback_add(handle, &new_entry->efd_jce);
+	sbi->s_mb_free_pending += clusters;
+	spin_unlock(&sbi->s_md_lock);
 	return 0;
 }
 
-- 
2.5.0

Re: [PATCH -v2] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Andreas Dilger]

[-- Attachment #1: Type: text/plain, Size: 5108 bytes --]

On Jun 7, 2016, at 8:46 PM, Theodore Ts'o <tytso@mit.edu> wrote:
> 
> If there are pending blocks to be released after a commit, retrying
> the allocation after a journal commit has no hope of helping.  So
> track how many pending deleted blocks there might be, and don't retry
> if there are no pending blocks.
> 
> Reported-by: Chao Yu <yuchao0@huawei.com>
> Signed-off-by: Theodore Ts'o <tytso@mit.edu>
> ---
> 
> Oops, ignore the earlier version of this patch.  I bobbled the commit
> and merged in part of another change.
> 
> fs/ext4/balloc.c    |  9 ++++++++-
> fs/ext4/ext4.h      |  1 +
> fs/ext4/ext4_jbd2.h | 10 +++++++++-
> fs/ext4/mballoc.c   | 12 ++++++++++--
> 4 files changed, 28 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
> index 3020fd7..371ac63 100644
> --- a/fs/ext4/balloc.c
> +++ b/fs/ext4/balloc.c
> @@ -603,7 +603,14 @@ int ext4_claim_free_clusters(struct ext4_sb_info *sbi,
>  */
> int ext4_should_retry_alloc(struct super_block *sb, int *retries)
> {
> -	if (!ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
> +	unsigned int pending_blocks;
> +
> +	spin_lock(&EXT4_SB(sb)->s_md_lock);
> +	pending_blocks = EXT4_SB(sb)->s_mb_free_pending;
> +	spin_unlock(&EXT4_SB(sb)->s_md_lock);

Is there a benefit of having a spinlock here?  The read of free_pending
should be atomic.

Cheers, Andreas

> +
> +	if (pending_blocks == 0 ||
> +	    !ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
> 	    (*retries)++ > 3 ||
> 	    !EXT4_SB(sb)->s_journal)
> 		return 0;
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index b84aa1c..96c73e6 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -1430,6 +1430,7 @@ struct ext4_sb_info {
> 	unsigned short *s_mb_offsets;
> 	unsigned int *s_mb_maxs;
> 	unsigned int s_group_info_size;
> +	unsigned int s_mb_free_pending;
> 
> 	/* tunables */
> 	unsigned long s_stripe;
> diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
> index 09c1ef3..b1d52c1 100644
> --- a/fs/ext4/ext4_jbd2.h
> +++ b/fs/ext4/ext4_jbd2.h
> @@ -175,6 +175,13 @@ struct ext4_journal_cb_entry {
>  * There is no guaranteed calling order of multiple registered callbacks on
>  * the same transaction.
>  */
> +static inline void _ext4_journal_callback_add(handle_t *handle,
> +			struct ext4_journal_cb_entry *jce)
> +{
> +	/* Add the jce to transaction's private list */
> +	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
> +}
> +
> static inline void ext4_journal_callback_add(handle_t *handle,
> 			void (*func)(struct super_block *sb,
> 				     struct ext4_journal_cb_entry *jce,
> @@ -187,10 +194,11 @@ static inline void ext4_journal_callback_add(handle_t *handle,
> 	/* Add the jce to transaction's private list */
> 	jce->jce_func = func;
> 	spin_lock(&sbi->s_md_lock);
> -	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
> +	_ext4_journal_callback_add(handle, jce);
> 	spin_unlock(&sbi->s_md_lock);
> }
> 
> +
> /**
>  * ext4_journal_callback_del: delete a registered callback
>  * @handle: active journal transaction handle on which callback was registered
> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> index c1ab3ec..77249e1 100644
> --- a/fs/ext4/mballoc.c
> +++ b/fs/ext4/mballoc.c
> @@ -2627,6 +2627,7 @@ int ext4_mb_init(struct super_block *sb)
> 
> 	spin_lock_init(&sbi->s_md_lock);
> 	spin_lock_init(&sbi->s_bal_lock);
> +	sbi->s_mb_free_pending = 0;
> 
> 	sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN;
> 	sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
> @@ -2814,6 +2815,9 @@ static void ext4_free_data_callback(struct super_block *sb,
> 	/* we expect to find existing buddy because it's pinned */
> 	BUG_ON(err != 0);
> 
> +	spin_lock(&EXT4_SB(sb)->s_md_lock);
> +	EXT4_SB(sb)->s_mb_free_pending -= entry->efd_count;
> +	spin_unlock(&EXT4_SB(sb)->s_md_lock);
> 
> 	db = e4b.bd_info;
> 	/* there are blocks to put in buddy to make them really free */
> @@ -4583,6 +4587,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
> {
> 	ext4_group_t group = e4b->bd_group;
> 	ext4_grpblk_t cluster;
> +	ext4_grpblk_t clusters = new_entry->efd_count;
> 	struct ext4_free_data *entry;
> 	struct ext4_group_info *db = e4b->bd_info;
> 	struct super_block *sb = e4b->bd_sb;
> @@ -4649,8 +4654,11 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
> 		}
> 	}
> 	/* Add the extent to transaction's private list */
> -	ext4_journal_callback_add(handle, ext4_free_data_callback,
> -				  &new_entry->efd_jce);
> +	new_entry->efd_jce.jce_func = ext4_free_data_callback;
> +	spin_lock(&sbi->s_md_lock);
> +	_ext4_journal_callback_add(handle, &new_entry->efd_jce);
> +	sbi->s_mb_free_pending += clusters;
> +	spin_unlock(&sbi->s_md_lock);
> 	return 0;
> }
> 
> --
> 2.5.0
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


Cheers, Andreas






[-- Attachment #2: Message signed with OpenPGP using GPGMail --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH -v2] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Jan Kara]

On Tue 07-06-16 22:46:46, Ted Tso wrote:
> If there are pending blocks to be released after a commit, retrying
> the allocation after a journal commit has no hope of helping.  So
> track how many pending deleted blocks there might be, and don't retry
> if there are no pending blocks.
> 
> Reported-by: Chao Yu <yuchao0@huawei.com>
> Signed-off-by: Theodore Ts'o <tytso@mit.edu>
> ---
> 
> Oops, ignore the earlier version of this patch.  I bobbled the commit
> and merged in part of another change.

Couple of notes below:

>  fs/ext4/balloc.c    |  9 ++++++++-
>  fs/ext4/ext4.h      |  1 +
>  fs/ext4/ext4_jbd2.h | 10 +++++++++-
>  fs/ext4/mballoc.c   | 12 ++++++++++--
>  4 files changed, 28 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
> index 3020fd7..371ac63 100644
> --- a/fs/ext4/balloc.c
> +++ b/fs/ext4/balloc.c
> @@ -603,7 +603,14 @@ int ext4_claim_free_clusters(struct ext4_sb_info *sbi,
>   */
>  int ext4_should_retry_alloc(struct super_block *sb, int *retries)
>  {
> -	if (!ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
> +	unsigned int pending_blocks;
> +
> +	spin_lock(&EXT4_SB(sb)->s_md_lock);
> +	pending_blocks = EXT4_SB(sb)->s_mb_free_pending;
> +	spin_unlock(&EXT4_SB(sb)->s_md_lock);
> +
> +	if (pending_blocks == 0 ||
> +	    !ext4_has_free_clusters(EXT4_SB(sb), 1, 0) ||
>  	    (*retries)++ > 3 ||
>  	    !EXT4_SB(sb)->s_journal)
>  		return 0;

But this is racy. Transaction commit could have finished before we called
ext4_should_retry_alloc() and so we will mistakenly think there's no hope
although there are blocks free now. But what you could probably do is just
return 1 without forcing a transaction commit when pending_blocks == 0.

> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index b84aa1c..96c73e6 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -1430,6 +1430,7 @@ struct ext4_sb_info {
>  	unsigned short *s_mb_offsets;
>  	unsigned int *s_mb_maxs;
>  	unsigned int s_group_info_size;
> +	unsigned int s_mb_free_pending;
>  
>  	/* tunables */
>  	unsigned long s_stripe;
> diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
> index 09c1ef3..b1d52c1 100644
> --- a/fs/ext4/ext4_jbd2.h
> +++ b/fs/ext4/ext4_jbd2.h
> @@ -175,6 +175,13 @@ struct ext4_journal_cb_entry {
>   * There is no guaranteed calling order of multiple registered callbacks on
>   * the same transaction.
>   */
> +static inline void _ext4_journal_callback_add(handle_t *handle,
> +			struct ext4_journal_cb_entry *jce)
> +{
> +	/* Add the jce to transaction's private list */
> +	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
> +}
> +
>  static inline void ext4_journal_callback_add(handle_t *handle,
>  			void (*func)(struct super_block *sb,
>  				     struct ext4_journal_cb_entry *jce,

Well, since ext4_mb_free_metadata() is the only user of
ext4_journal_callback_add(), ext4_journal_callback_add() won't have any
user after your patch. Maybe we could just stop playing these abstraction
games nobody currently uses and just implement a helper function to add
freeing callback to the transaction list including increment of the pending
counter.

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: [PATCH -v2] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Andreas Dilger]

[-- Attachment #1: Type: text/plain, Size: 530 bytes --]

On Jun 16, 2016, at 1:20 PM, Jan Kara <jack@suse.cz> wrote:
> 
> Well, since ext4_mb_free_metadata() is the only user of
> ext4_journal_callback_add(), ext4_journal_callback_add() won't have any
> user after your patch. Maybe we could just stop playing these abstraction
> games nobody currently uses and just implement a helper function to add
> freeing callback to the transaction list including increment of the pending
> counter.

We use the journal callback for Lustre servers.  Please don't remove it.

Cheers, Andreas






[-- Attachment #2: Message signed with OpenPGP using GPGMail --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH -v2] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Jan Kara]

On Thu 16-06-16 14:11:28, Andreas Dilger wrote:
> On Jun 16, 2016, at 1:20 PM, Jan Kara <jack@suse.cz> wrote:
> > 
> > Well, since ext4_mb_free_metadata() is the only user of
> > ext4_journal_callback_add(), ext4_journal_callback_add() won't have any
> > user after your patch. Maybe we could just stop playing these abstraction
> > games nobody currently uses and just implement a helper function to add
> > freeing callback to the transaction list including increment of the pending
> > counter.
> 
> We use the journal callback for Lustre servers.  Please don't remove it.

I didn't really mean to rewrite the callback infrastructure. I just meant
to change the helper function to fit in-tree users. If Lustre needs to do
something else, it is free to have its own helper function which attaches
callbacks to the transaction's list.

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

[PATCH -v3] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Theodore Ts'o]

If there are no pending blocks to be released after a commit, forcing
a journal commit has no hope of helping.  It's possible that a commit
had just completed, so if there are now free blocks available for
allocation, it's worth retrying the commit.

Reported-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/balloc.c    |  4 +++-
 fs/ext4/ext4.h      |  1 +
 fs/ext4/ext4_jbd2.h | 10 +++++++++-
 fs/ext4/mballoc.c   | 12 ++++++++++--
 4 files changed, 23 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 3020fd7..0b8105b 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -610,7 +610,9 @@ int ext4_should_retry_alloc(struct super_block *sb, int *retries)
 
 	jbd_debug(1, "%s: retrying operation after ENOSPC\n", sb->s_id);
 
-	jbd2_journal_force_commit_nested(EXT4_SB(sb)->s_journal);
+	smp_mb();
+	if (EXT4_SB(sb)->s_mb_free_pending)
+		jbd2_journal_force_commit_nested(EXT4_SB(sb)->s_journal);
 	return 1;
 }
 
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index b84aa1c..96c73e6 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1430,6 +1430,7 @@ struct ext4_sb_info {
 	unsigned short *s_mb_offsets;
 	unsigned int *s_mb_maxs;
 	unsigned int s_group_info_size;
+	unsigned int s_mb_free_pending;
 
 	/* tunables */
 	unsigned long s_stripe;
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index 09c1ef3..b1d52c1 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -175,6 +175,13 @@ struct ext4_journal_cb_entry {
  * There is no guaranteed calling order of multiple registered callbacks on
  * the same transaction.
  */
+static inline void _ext4_journal_callback_add(handle_t *handle,
+			struct ext4_journal_cb_entry *jce)
+{
+	/* Add the jce to transaction's private list */
+	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
+}
+
 static inline void ext4_journal_callback_add(handle_t *handle,
 			void (*func)(struct super_block *sb,
 				     struct ext4_journal_cb_entry *jce,
@@ -187,10 +194,11 @@ static inline void ext4_journal_callback_add(handle_t *handle,
 	/* Add the jce to transaction's private list */
 	jce->jce_func = func;
 	spin_lock(&sbi->s_md_lock);
-	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
+	_ext4_journal_callback_add(handle, jce);
 	spin_unlock(&sbi->s_md_lock);
 }
 
+
 /**
  * ext4_journal_callback_del: delete a registered callback
  * @handle: active journal transaction handle on which callback was registered
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index c1ab3ec..77249e1 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2627,6 +2627,7 @@ int ext4_mb_init(struct super_block *sb)
 
 	spin_lock_init(&sbi->s_md_lock);
 	spin_lock_init(&sbi->s_bal_lock);
+	sbi->s_mb_free_pending = 0;
 
 	sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN;
 	sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
@@ -2814,6 +2815,9 @@ static void ext4_free_data_callback(struct super_block *sb,
 	/* we expect to find existing buddy because it's pinned */
 	BUG_ON(err != 0);
 
+	spin_lock(&EXT4_SB(sb)->s_md_lock);
+	EXT4_SB(sb)->s_mb_free_pending -= entry->efd_count;
+	spin_unlock(&EXT4_SB(sb)->s_md_lock);
 
 	db = e4b.bd_info;
 	/* there are blocks to put in buddy to make them really free */
@@ -4583,6 +4587,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 {
 	ext4_group_t group = e4b->bd_group;
 	ext4_grpblk_t cluster;
+	ext4_grpblk_t clusters = new_entry->efd_count;
 	struct ext4_free_data *entry;
 	struct ext4_group_info *db = e4b->bd_info;
 	struct super_block *sb = e4b->bd_sb;
@@ -4649,8 +4654,11 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
 		}
 	}
 	/* Add the extent to transaction's private list */
-	ext4_journal_callback_add(handle, ext4_free_data_callback,
-				  &new_entry->efd_jce);
+	new_entry->efd_jce.jce_func = ext4_free_data_callback;
+	spin_lock(&sbi->s_md_lock);
+	_ext4_journal_callback_add(handle, &new_entry->efd_jce);
+	sbi->s_mb_free_pending += clusters;
+	spin_unlock(&sbi->s_md_lock);
 	return 0;
 }
 
-- 
2.5.0

Re: [PATCH -v3] ext4: optimize ext4_should_retry_alloc() to improve ENOSPC performance

[Jan Kara]

On Thu 30-06-16 11:00:31, Ted Tso wrote:
> If there are no pending blocks to be released after a commit, forcing
> a journal commit has no hope of helping.  It's possible that a commit
> had just completed, so if there are now free blocks available for
> allocation, it's worth retrying the commit.
> 
> Reported-by: Chao Yu <yuchao0@huawei.com>
> Signed-off-by: Theodore Ts'o <tytso@mit.edu>

The patch looks good now. Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

I'd just remove ext4_journal_callback_add() function since it is unused now
and if Lustre uses it, it can easily replicate it.

								Honza
> ---
>  fs/ext4/balloc.c    |  4 +++-
>  fs/ext4/ext4.h      |  1 +
>  fs/ext4/ext4_jbd2.h | 10 +++++++++-
>  fs/ext4/mballoc.c   | 12 ++++++++++--
>  4 files changed, 23 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
> index 3020fd7..0b8105b 100644
> --- a/fs/ext4/balloc.c
> +++ b/fs/ext4/balloc.c
> @@ -610,7 +610,9 @@ int ext4_should_retry_alloc(struct super_block *sb, int *retries)
>  
>  	jbd_debug(1, "%s: retrying operation after ENOSPC\n", sb->s_id);
>  
> -	jbd2_journal_force_commit_nested(EXT4_SB(sb)->s_journal);
> +	smp_mb();
> +	if (EXT4_SB(sb)->s_mb_free_pending)
> +		jbd2_journal_force_commit_nested(EXT4_SB(sb)->s_journal);
>  	return 1;
>  }
>  
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index b84aa1c..96c73e6 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -1430,6 +1430,7 @@ struct ext4_sb_info {
>  	unsigned short *s_mb_offsets;
>  	unsigned int *s_mb_maxs;
>  	unsigned int s_group_info_size;
> +	unsigned int s_mb_free_pending;
>  
>  	/* tunables */
>  	unsigned long s_stripe;
> diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
> index 09c1ef3..b1d52c1 100644
> --- a/fs/ext4/ext4_jbd2.h
> +++ b/fs/ext4/ext4_jbd2.h
> @@ -175,6 +175,13 @@ struct ext4_journal_cb_entry {
>   * There is no guaranteed calling order of multiple registered callbacks on
>   * the same transaction.
>   */
> +static inline void _ext4_journal_callback_add(handle_t *handle,
> +			struct ext4_journal_cb_entry *jce)
> +{
> +	/* Add the jce to transaction's private list */
> +	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
> +}
> +
>  static inline void ext4_journal_callback_add(handle_t *handle,
>  			void (*func)(struct super_block *sb,
>  				     struct ext4_journal_cb_entry *jce,
> @@ -187,10 +194,11 @@ static inline void ext4_journal_callback_add(handle_t *handle,
>  	/* Add the jce to transaction's private list */
>  	jce->jce_func = func;
>  	spin_lock(&sbi->s_md_lock);
> -	list_add_tail(&jce->jce_list, &handle->h_transaction->t_private_list);
> +	_ext4_journal_callback_add(handle, jce);
>  	spin_unlock(&sbi->s_md_lock);
>  }
>  
> +
>  /**
>   * ext4_journal_callback_del: delete a registered callback
>   * @handle: active journal transaction handle on which callback was registered
> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> index c1ab3ec..77249e1 100644
> --- a/fs/ext4/mballoc.c
> +++ b/fs/ext4/mballoc.c
> @@ -2627,6 +2627,7 @@ int ext4_mb_init(struct super_block *sb)
>  
>  	spin_lock_init(&sbi->s_md_lock);
>  	spin_lock_init(&sbi->s_bal_lock);
> +	sbi->s_mb_free_pending = 0;
>  
>  	sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN;
>  	sbi->s_mb_min_to_scan = MB_DEFAULT_MIN_TO_SCAN;
> @@ -2814,6 +2815,9 @@ static void ext4_free_data_callback(struct super_block *sb,
>  	/* we expect to find existing buddy because it's pinned */
>  	BUG_ON(err != 0);
>  
> +	spin_lock(&EXT4_SB(sb)->s_md_lock);
> +	EXT4_SB(sb)->s_mb_free_pending -= entry->efd_count;
> +	spin_unlock(&EXT4_SB(sb)->s_md_lock);
>  
>  	db = e4b.bd_info;
>  	/* there are blocks to put in buddy to make them really free */
> @@ -4583,6 +4587,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
>  {
>  	ext4_group_t group = e4b->bd_group;
>  	ext4_grpblk_t cluster;
> +	ext4_grpblk_t clusters = new_entry->efd_count;
>  	struct ext4_free_data *entry;
>  	struct ext4_group_info *db = e4b->bd_info;
>  	struct super_block *sb = e4b->bd_sb;
> @@ -4649,8 +4654,11 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
>  		}
>  	}
>  	/* Add the extent to transaction's private list */
> -	ext4_journal_callback_add(handle, ext4_free_data_callback,
> -				  &new_entry->efd_jce);
> +	new_entry->efd_jce.jce_func = ext4_free_data_callback;
> +	spin_lock(&sbi->s_md_lock);
> +	_ext4_journal_callback_add(handle, &new_entry->efd_jce);
> +	sbi->s_mb_free_pending += clusters;
> +	spin_unlock(&sbi->s_md_lock);
>  	return 0;
>  }
>  
> -- 
> 2.5.0
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR