From: Miao Xie <miaox@cn.fujitsu.com>
To: Linux Btrfs <linux-btrfs@vger.kernel.org>
Subject: [PATCH V4 04/12] Btrfs: use a slab for ordered extents allocation
Date: Thu, 06 Sep 2012 18:01:51 +0800 [thread overview]
Message-ID: <5048748F.2070601@cn.fujitsu.com> (raw)
In-Reply-To: <50487434.1010305@cn.fujitsu.com>
The ordered extent allocation is in the fast path of the IO, so use a slab
to improve the speed of the allocation.
"Size of the struct is 280, so this will fall into the size-512 bucket,
giving 8 objects per page, while own slab will pack 14 objects into a page.
Another benefit I see is to check for leaked objects when the module is
removed (and the cache destroy takes place)."
-- David Sterba
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
---
Changelog v1 -> v4:
- just change the changelog of this patch.
---
fs/btrfs/ordered-data.c | 23 +++++++++++++++++++++--
fs/btrfs/ordered-data.h | 2 ++
fs/btrfs/super.c | 9 ++++++++-
3 files changed, 31 insertions(+), 3 deletions(-)
diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c
index 2eb79cc..4ae1014 100644
--- a/fs/btrfs/ordered-data.c
+++ b/fs/btrfs/ordered-data.c
@@ -25,6 +25,8 @@
#include "btrfs_inode.h"
#include "extent_io.h"
+static struct kmem_cache *btrfs_ordered_extent_cache;
+
static u64 entry_end(struct btrfs_ordered_extent *entry)
{
if (entry->file_offset + entry->len < entry->file_offset)
@@ -187,7 +189,7 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset,
struct btrfs_ordered_extent *entry;
tree = &BTRFS_I(inode)->ordered_tree;
- entry = kzalloc(sizeof(*entry), GFP_NOFS);
+ entry = kmem_cache_zalloc(btrfs_ordered_extent_cache, GFP_NOFS);
if (!entry)
return -ENOMEM;
@@ -421,7 +423,7 @@ void btrfs_put_ordered_extent(struct btrfs_ordered_extent *entry)
list_del(&sum->list);
kfree(sum);
}
- kfree(entry);
+ kmem_cache_free(btrfs_ordered_extent_cache, entry);
}
}
@@ -958,3 +960,20 @@ void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
}
spin_unlock(&root->fs_info->ordered_extent_lock);
}
+
+int __init ordered_data_init(void)
+{
+ btrfs_ordered_extent_cache = kmem_cache_create("btrfs_ordered_extent",
+ sizeof(struct btrfs_ordered_extent), 0,
+ SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD,
+ NULL);
+ if (!btrfs_ordered_extent_cache)
+ return -ENOMEM;
+ return 0;
+}
+
+void ordered_data_exit(void)
+{
+ if (btrfs_ordered_extent_cache)
+ kmem_cache_destroy(btrfs_ordered_extent_cache);
+}
diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h
index c2443a4..d1ddaef 100644
--- a/fs/btrfs/ordered-data.h
+++ b/fs/btrfs/ordered-data.h
@@ -192,4 +192,6 @@ void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
struct inode *inode);
void btrfs_wait_ordered_extents(struct btrfs_root *root,
int nocow_only, int delay_iput);
+int __init ordered_data_init(void);
+void ordered_data_exit(void);
#endif
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 073c236..48e53d6 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -1595,10 +1595,14 @@ static int __init init_btrfs_fs(void)
if (err)
goto free_extent_io;
- err = btrfs_delayed_inode_init();
+ err = ordered_data_init();
if (err)
goto free_extent_map;
+ err = btrfs_delayed_inode_init();
+ if (err)
+ goto free_ordered_data;
+
err = btrfs_interface_init();
if (err)
goto free_delayed_inode;
@@ -1616,6 +1620,8 @@ unregister_ioctl:
btrfs_interface_exit();
free_delayed_inode:
btrfs_delayed_inode_exit();
+free_ordered_data:
+ ordered_data_exit();
free_extent_map:
extent_map_exit();
free_extent_io:
@@ -1632,6 +1638,7 @@ static void __exit exit_btrfs_fs(void)
{
btrfs_destroy_cachep();
btrfs_delayed_inode_exit();
+ ordered_data_exit();
extent_map_exit();
extent_io_exit();
btrfs_interface_exit();
--
1.7.6.5
next prev parent reply other threads:[~2012-09-06 10:44 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-09-06 10:00 [PATCH V4 0/12] random bug fixes Miao Xie
2012-09-06 10:00 ` [PATCH V4 01/12] Btrfs: fix error path in create_pending_snapshot() Miao Xie
2012-09-17 16:56 ` David Sterba
2012-09-18 1:47 ` Miao Xie
2012-09-06 10:00 ` [PATCH V4 02/12] Btrfs: fix full backref problem when inserting shared block reference Miao Xie
2012-09-06 10:01 ` [PATCH V4 03/12] Btrfs: fix file extent discount problem in the, snapshot Miao Xie
2012-09-06 10:01 ` Miao Xie [this message]
2012-09-06 10:02 ` [PATCH V4 05/12] Btrfs: fix wrong orphan count of the fs/file tree Miao Xie
2012-09-06 10:02 ` [PATCH V4 06/12] Btrfs: add a new "type" field into the block reservation structure Miao Xie
2012-09-06 10:03 ` [PATCH V4 07/12] Btrfs: fix corrupted metadata in the snapshot Miao Xie
2012-09-06 13:09 ` Josef Bacik
2012-09-07 3:10 ` Miao Xie
2012-09-07 7:43 ` [PATCH V5 " Miao Xie
2012-09-06 10:03 ` [PATCH V4 08/12] Btrfs: fix the snapshot that should not exist Miao Xie
2012-09-06 10:03 ` [PATCH V4 09/12] Btrfs: fix wrong size for the reservation of the, snapshot creation Miao Xie
2012-09-06 10:04 ` [PATCH V4 10/12] Btrfs: fix unprotected ->log_batch Miao Xie
2012-09-06 10:04 ` [PATCH V4 11/12] Btrfs: output more information when aborting a unused transaction handle Miao Xie
2012-09-06 10:04 ` [PATCH V4 12/12] Btrfs: fix wrong size for the reservation when doing, file pre-allocation Miao Xie
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=5048748F.2070601@cn.fujitsu.com \
--to=miaox@cn.fujitsu.com \
--cc=linux-btrfs@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).