From: Jan Kara <jack@suse.cz>
To: Ted Tso <tytso@mit.edu>
Cc: linux-ext4@vger.kernel.org, Jan Kara <jack@suse.cz>
Subject: [PATCH 14/26] ext4: Better estimate credits needed for ext4_da_writepages()
Date: Fri, 31 May 2013 11:42:47 +0200 [thread overview]
Message-ID: <1369993379-13017-15-git-send-email-jack@suse.cz> (raw)
In-Reply-To: <1369993379-13017-1-git-send-email-jack@suse.cz>
We limit the number of blocks written in a single loop of
ext4_da_writepages() to 64 when inode uses indirect blocks. That is
unnecessary as credit estimates for mapping logically continguous run of
blocks is rather low even for inode with indirect blocks. So just lift
this limitation and properly calculate the number of necessary credits.
This better credit estimate will also later allow us to always write at
least a single page in one iteration.
Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>
Signed-off-by: Jan Kara <jack@suse.cz>
---
fs/ext4/ext4.h | 3 +--
fs/ext4/extents.c | 16 +++++++-------
fs/ext4/inode.c | 63 +++++++++++++++++++++++++------------------------------
3 files changed, 36 insertions(+), 46 deletions(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index bca9889..4ed13c4 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2598,8 +2598,7 @@ struct ext4_extent;
extern int ext4_ext_tree_init(handle_t *handle, struct inode *);
extern int ext4_ext_writepage_trans_blocks(struct inode *, int);
-extern int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks,
- int chunk);
+extern int ext4_ext_index_trans_blocks(struct inode *inode, int extents);
extern int ext4_ext_map_blocks(handle_t *handle, struct inode *inode,
struct ext4_map_blocks *map, int flags);
extern void ext4_ext_truncate(handle_t *, struct inode *);
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index bc0f191..e455141 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2328,17 +2328,15 @@ int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks,
}
/*
- * How many index/leaf blocks need to change/allocate to modify nrblocks?
+ * How many index/leaf blocks need to change/allocate to add @extents extents?
*
- * if nrblocks are fit in a single extent (chunk flag is 1), then
- * in the worse case, each tree level index/leaf need to be changed
- * if the tree split due to insert a new extent, then the old tree
- * index/leaf need to be updated too
+ * If we add a single extent, then in the worse case, each tree level
+ * index/leaf need to be changed in case of the tree split.
*
- * If the nrblocks are discontiguous, they could cause
- * the whole tree split more than once, but this is really rare.
+ * If more extents are inserted, they could cause the whole tree split more
+ * than once, but this is really rare.
*/
-int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
+int ext4_ext_index_trans_blocks(struct inode *inode, int extents)
{
int index;
int depth;
@@ -2349,7 +2347,7 @@ int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
depth = ext_depth(inode);
- if (chunk)
+ if (extents <= 1)
index = depth * 2;
else
index = depth * 3;
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 27b8504..db33a00 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -138,6 +138,9 @@ static int ext4_bh_delay_or_unwritten(handle_t *handle, struct buffer_head *bh);
static int ext4_discard_partial_page_buffers_no_lock(handle_t *handle,
struct inode *inode, struct page *page, loff_t from,
loff_t length, int flags);
+static int ext4_meta_trans_blocks(struct inode *inode, int lblocks,
+ int pextents);
+
/*
* Test whether an inode is a fast symlink.
@@ -2198,28 +2201,25 @@ static int ext4_writepage(struct page *page,
}
/*
- * This is called via ext4_da_writepages() to
- * calculate the total number of credits to reserve to fit
- * a single extent allocation into a single transaction,
- * ext4_da_writpeages() will loop calling this before
- * the block allocation.
+ * mballoc gives us at most this number of blocks...
+ * XXX: That seems to be only a limitation of ext4_mb_normalize_request().
+ * The rest of mballoc seems to handle chunks upto full group size.
*/
+#define MAX_WRITEPAGES_EXTENT_LEN 2048
+/*
+ * Calculate the total number of credits to reserve for one writepages
+ * iteration. This is called from ext4_da_writepages(). We map an extent of
+ * upto MAX_WRITEPAGES_EXTENT_LEN blocks and then we go on and finish mapping
+ * the last partial page. So in total we can map MAX_WRITEPAGES_EXTENT_LEN +
+ * bpp - 1 blocks in bpp different extents.
+ */
static int ext4_da_writepages_trans_blocks(struct inode *inode)
{
- int max_blocks = EXT4_I(inode)->i_reserved_data_blocks;
-
- /*
- * With non-extent format the journal credit needed to
- * insert nrblocks contiguous block is dependent on
- * number of contiguous block. So we will limit
- * number of contiguous block to a sane value
- */
- if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) &&
- (max_blocks > EXT4_MAX_TRANS_DATA))
- max_blocks = EXT4_MAX_TRANS_DATA;
+ int bpp = ext4_journal_blocks_per_page(inode);
- return ext4_chunk_trans_blocks(inode, max_blocks);
+ return ext4_meta_trans_blocks(inode,
+ MAX_WRITEPAGES_EXTENT_LEN + bpp - 1, bpp);
}
/*
@@ -4753,11 +4753,12 @@ int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry,
return 0;
}
-static int ext4_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
+static int ext4_index_trans_blocks(struct inode *inode, int lblocks,
+ int pextents)
{
if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
- return ext4_ind_trans_blocks(inode, nrblocks);
- return ext4_ext_index_trans_blocks(inode, nrblocks, chunk);
+ return ext4_ind_trans_blocks(inode, lblocks);
+ return ext4_ext_index_trans_blocks(inode, pextents);
}
/*
@@ -4771,7 +4772,8 @@ static int ext4_index_trans_blocks(struct inode *inode, int nrblocks, int chunk)
*
* Also account for superblock, inode, quota and xattr blocks
*/
-static int ext4_meta_trans_blocks(struct inode *inode, int nrblocks, int chunk)
+static int ext4_meta_trans_blocks(struct inode *inode, int lblocks,
+ int pextents)
{
ext4_group_t groups, ngroups = ext4_get_groups_count(inode->i_sb);
int gdpblocks;
@@ -4779,14 +4781,10 @@ static int ext4_meta_trans_blocks(struct inode *inode, int nrblocks, int chunk)
int ret = 0;
/*
- * How many index blocks need to touch to modify nrblocks?
- * The "Chunk" flag indicating whether the nrblocks is
- * physically contiguous on disk
- *
- * For Direct IO and fallocate, they calls get_block to allocate
- * one single extent at a time, so they could set the "Chunk" flag
+ * How many index blocks need to touch to map @lblocks logical blocks
+ * to @pextents physical extents?
*/
- idxblocks = ext4_index_trans_blocks(inode, nrblocks, chunk);
+ idxblocks = ext4_index_trans_blocks(inode, lblocks, pextents);
ret = idxblocks;
@@ -4794,12 +4792,7 @@ static int ext4_meta_trans_blocks(struct inode *inode, int nrblocks, int chunk)
* Now let's see how many group bitmaps and group descriptors need
* to account
*/
- groups = idxblocks;
- if (chunk)
- groups += 1;
- else
- groups += nrblocks;
next prev parent reply other threads:[~2013-05-31 9:43 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-05-31 9:42 [PATCH 00/22 v2] Fixes and improvements in ext4 writeback path Jan Kara
2013-05-31 9:42 ` [PATCH 01/26] ext4: use io_end for multiple bios Jan Kara
2013-06-04 16:03 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 02/26] jbd2: Don't create journal_head for temporary journal buffers Jan Kara
2013-06-04 16:04 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 03/26] jbd2: Remove journal_head from descriptor buffers Jan Kara
2013-06-04 16:06 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 04/26] jbd2: Refine waiting for shadow buffers Jan Kara
2013-06-04 16:09 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 05/26] jbd2: Remove outdated comment Jan Kara
2013-06-04 16:11 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 06/26] jbd2: Cleanup needed free block estimates when starting a transaction Jan Kara
2013-06-04 16:17 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 07/26] jbd2: Fix race in t_outstanding_credits update in jbd2_journal_extend() Jan Kara
2013-06-04 16:23 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 08/26] jbd2: Remove unused waitqueues Jan Kara
2013-06-04 16:24 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 09/26] jbd2: Transaction reservation support Jan Kara
2013-06-04 16:36 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 10/26] ext4: Provide wrappers for transaction reservation calls Jan Kara
2013-06-04 16:41 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 11/26] ext4: Stop messing with nr_to_write in ext4_da_writepages() Jan Kara
2013-06-04 16:49 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 12/26] ext4: Deprecate max_writeback_mb_bump sysfs attribute Jan Kara
2013-06-04 16:53 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 13/26] ext4: Improve writepage credit estimate for files with indirect blocks Jan Kara
2013-06-03 21:45 ` Darrick J. Wong
2013-06-04 16:57 ` Theodore Ts'o
2013-05-31 9:42 ` Jan Kara [this message]
2013-06-04 17:01 ` [PATCH 14/26] ext4: Better estimate credits needed for ext4_da_writepages() Theodore Ts'o
2013-05-31 9:42 ` [PATCH 15/26] ext4: Restructure writeback path Jan Kara
2013-06-04 17:18 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 16/26] ext4: Remove buffer_uninit handling Jan Kara
2013-06-04 17:20 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 17/26] ext4: Use transaction reservation for extent conversion in ext4_end_io Jan Kara
2013-06-04 17:29 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 18/26] ext4: Split extent conversion lists to reserved & unreserved parts Jan Kara
2013-06-04 18:22 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 19/26] ext4: Defer clearing of PageWriteback after extent conversion Jan Kara
2013-06-04 18:24 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 20/26] ext4: Protect extent conversion after DIO with i_dio_count Jan Kara
2013-06-04 18:28 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 21/26] ext4: Remove wait for unwritten extent conversion from ext4_truncate() Jan Kara
2013-06-04 18:30 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 22/26] ext4: Use generic_file_fsync() in ext4_file_fsync() in nojournal mode Jan Kara
2013-06-04 18:38 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 23/26] ext4: Remove i_mutex from ext4_file_sync() Jan Kara
2013-06-04 18:40 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 24/26] ext4: Remove wait for unwritten extents in ext4_ind_direct_IO() Jan Kara
2013-06-04 18:42 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 25/26] ext4: Don't wait for extent conversion in ext4_punch_hole() Jan Kara
2013-06-04 18:46 ` Theodore Ts'o
2013-05-31 9:42 ` [PATCH 26/26] ext4: Remove ext4_ioend_wait() Jan Kara
2013-06-04 18:47 ` Theodore Ts'o
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=1369993379-13017-15-git-send-email-jack@suse.cz \
--to=jack@suse.cz \
--cc=linux-ext4@vger.kernel.org \
--cc=tytso@mit.edu \
/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).