From: "Theodore Ts'o" <tytso@mit.edu>
To: Linux Kernel Developers List <linux-kernel@vger.kernel.org>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Subject: [PATCH 35/49] ext4: Clean up ext4_get_blocks() so it does not depend on bh_result->b_state
Date: Mon, 8 Jun 2009 15:22:53 -0400 [thread overview]
Message-ID: <1244488987-32564-36-git-send-email-tytso@mit.edu> (raw)
In-Reply-To: <1244488987-32564-35-git-send-email-tytso@mit.edu>
The ext4_get_blocks() function was depending on the value of
bh_result->b_state as an input parameter to decide whether or not
update the delalloc accounting statistics by calling
ext4_da_update_reserve_space(). We now use a separate flag,
EXT4_GET_BLOCKS_UPDATE_RESERVE_SPACE, to requests this update, so that
all callers of ext4_get_blocks() can clear map_bh.b_state before
calling ext4_get_blocks() without worrying about any consistency
issues.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
fs/ext4/ext4.h | 15 +++++++++----
fs/ext4/inode.c | 56 ++++++++++++++++++++++++++++++------------------------
2 files changed, 41 insertions(+), 30 deletions(-)
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 17feb4a..d164f12 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -318,16 +318,21 @@ struct ext4_new_group_data {
*/
/* Allocate any needed blocks and/or convert an unitialized
extent to be an initialized ext4 */
-#define EXT4_GET_BLOCKS_CREATE 1
+#define EXT4_GET_BLOCKS_CREATE 0x0001
/* Request the creation of an unitialized extent */
-#define EXT4_GET_BLOCKS_UNINIT_EXT 2
+#define EXT4_GET_BLOCKS_UNINIT_EXT 0x0002
#define EXT4_GET_BLOCKS_CREATE_UNINIT_EXT (EXT4_GET_BLOCKS_UNINIT_EXT|\
EXT4_GET_BLOCKS_CREATE)
/* Update the ext4_inode_info i_disksize field */
-#define EXT4_GET_BLOCKS_EXTEND_DISKSIZE 4
+#define EXT4_GET_BLOCKS_EXTEND_DISKSIZE 0x0004
/* Caller is from the delayed allocation writeout path,
- so the filesystem blocks have already been accounted for */
-#define EXT4_GET_BLOCKS_DELALLOC_RESERVE 8
+ so set the magic i_delalloc_reserve_flag after taking the
+ inode allocation semaphore for */
+#define EXT4_GET_BLOCKS_DELALLOC_RESERVE 0x0008
+ /* Call ext4_da_update_reserve_space() after successfully
+ allocating the blocks */
+#define EXT4_GET_BLOCKS_UPDATE_RESERVE_SPACE 0x0010
+
/*
* ioctl commands
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index bfe50a2..d7b7480 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -1234,16 +1234,15 @@ int ext4_get_blocks(handle_t *handle, struct inode *inode, sector_t block,
}
}
- if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
+ if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
EXT4_I(inode)->i_delalloc_reserved_flag = 0;
- /*
- * Update reserved blocks/metadata blocks
- * after successful block allocation
- * which were deferred till now
- */
- if ((retval > 0) && buffer_delay(bh))
- ext4_da_update_reserve_space(inode, retval);
- }
+
+ /*
+ * Update reserved blocks/metadata blocks after successful
+ * block allocation which had been deferred till now.
+ */
+ if ((retval > 0) && (flags & EXT4_GET_BLOCKS_UPDATE_RESERVE_SPACE))
+ ext4_da_update_reserve_space(inode, retval);
up_write((&EXT4_I(inode)->i_data_sem));
return retval;
@@ -2015,7 +2014,7 @@ static void ext4_print_free_blocks(struct inode *inode)
*/
static int mpage_da_map_blocks(struct mpage_da_data *mpd)
{
- int err, blks;
+ int err, blks, get_blocks_flags;
struct buffer_head new;
sector_t next = mpd->b_blocknr;
unsigned max_blocks = mpd->b_size >> mpd->inode->i_blkbits;
@@ -2040,23 +2039,30 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd)
BUG_ON(!handle);
/*
- * We need to make sure the BH_Delay flag is passed down to
- * ext4_da_get_block_write(), since it calls ext4_get_blocks()
- * with the EXT4_GET_BLOCKS_DELALLOC_RESERVE flag. This flag
- * causes ext4_get_blocks() to call
- * ext4_da_update_reserve_space() if the passed buffer head
- * has the BH_Delay flag set. In the future, once we clean up
- * the interfaces to ext4_get_blocks(), we should pass in a
- * separate flag which requests that the delayed allocation
- * statistics should be updated, instead of depending on the
- * state information getting passed down via the map_bh's
- * state bitmasks plus the magic
- * EXT4_GET_BLOCKS_DELALLOC_RESERVE flag.
+ * Call ext4_get_blocks() to allocate any delayed allocation
+ * blocks, or to convert an uninitialized extent to be
+ * initialized (in the case where we have written into
+ * one or more preallocated blocks).
+ *
+ * We pass in the magic EXT4_GET_BLOCKS_DELALLOC_RESERVE to
+ * indicate that we are on the delayed allocation path. This
+ * affects functions in many different parts of the allocation
+ * call path. This flag exists primarily because we don't
+ * want to change *many* call functions, so ext4_get_blocks()
+ * will set the magic i_delalloc_reserved_flag once the
+ * inode's allocation semaphore is taken.
+ *
+ * If the blocks in questions were delalloc blocks, set
+ * EXT4_GET_BLOCKS_DELALLOC_RESERVE so the delalloc accounting
+ * variables are updated after the blocks have been allocated.
*/
- new.b_state = mpd->b_state & (1 << BH_Delay);
+ new.b_state = 0;
+ get_blocks_flags = (EXT4_GET_BLOCKS_CREATE |
+ EXT4_GET_BLOCKS_DELALLOC_RESERVE);
+ if (mpd->b_state & (1 << BH_Delay))
+ get_blocks_flags |= EXT4_GET_BLOCKS_UPDATE_RESERVE_SPACE;
blks = ext4_get_blocks(handle, mpd->inode, next, max_blocks,
- &new, EXT4_GET_BLOCKS_CREATE|
- EXT4_GET_BLOCKS_DELALLOC_RESERVE);
+ &new, get_blocks_flags);
if (blks < 0) {
err = blks;
/*
--
1.6.3.2.1.gb9f7d.dirty
next prev parent reply other threads:[~2009-06-08 19:28 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-06-08 19:22 [PATCH 00/49] Ext4 patches currently queued for mainline Theodore Ts'o
2009-06-08 19:22 ` [PATCH 01/49] ext4: Properly initialize the buffer_head state Theodore Ts'o
2009-06-08 19:22 ` [PATCH 02/49] vfs: Add BUG_ON for delayed and unwritten flags in submit_bh() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 03/49] ext4: Mark the unwritten buffer_head as mapped during write_begin Theodore Ts'o
2009-06-08 19:22 ` [PATCH 04/49] ext4: Fallback to vmalloc if kmalloc can't allocate s_flex_groups array Theodore Ts'o
2009-06-08 19:22 ` [PATCH 05/49] ext4: Use is_power_of_2() for clarity Theodore Ts'o
2009-06-08 19:22 ` [PATCH 06/49] ext3: avoid unnecessary spinlock in critical POSIX ACL path Theodore Ts'o
2009-06-08 19:22 ` [PATCH 07/49] ext4: " Theodore Ts'o
2009-06-08 19:22 ` [PATCH 08/49] ext4: Simplify ext4_commit_super()'s function signature Theodore Ts'o
2009-06-08 19:22 ` [PATCH 09/49] ext4: Fix and simplify s_dirt handling Theodore Ts'o
2009-06-08 19:22 ` [PATCH 10/49] ext4: Use separate super_operations structure for no_journal filesystems Theodore Ts'o
2009-06-08 19:22 ` [PATCH 11/49] ext4: Avoid races caused by on-line resizing and SMP memory reordering Theodore Ts'o
2009-06-08 19:22 ` [PATCH 12/49] ext4: Remove outdated comment about lock_super() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 13/49] ext4: ext4_mark_recovery_complete() doesn't need to use lock_super Theodore Ts'o
2009-06-08 19:22 ` [PATCH 14/49] ext4: Replace lock/unlock_super() with an explicit lock for the orphan list Theodore Ts'o
2009-06-08 19:22 ` [PATCH 15/49] ext4: Replace lock/unlock_super() with an explicit lock for resizing Theodore Ts'o
2009-06-08 19:22 ` [PATCH 16/49] ext4: Don't avoid using BLOCK_UNINIT block groups in mballoc Theodore Ts'o
2009-06-08 19:22 ` [PATCH 17/49] ext4: Move the ext4_i.h header file into ext4.h Theodore Ts'o
2009-06-08 19:22 ` [PATCH 18/49] ext4: Move the ext4_sb.h " Theodore Ts'o
2009-06-08 19:22 ` [PATCH 19/49] ext4: Move fs/ext4/namei.h " Theodore Ts'o
2009-06-08 19:22 ` [PATCH 20/49] ext4: Move fs/ext4/group.h " Theodore Ts'o
2009-06-08 19:22 ` [PATCH 21/49] ext4: Make the length of the mb_history file tunable Theodore Ts'o
2009-06-08 19:22 ` [PATCH 22/49] ext4: hook fiemap operation for directories Theodore Ts'o
2009-06-08 19:22 ` [PATCH 23/49] vfs: Enable FS_IOC_FIEMAP and FIGETBSZ for all filetypes Theodore Ts'o
2009-06-08 19:22 ` [PATCH 24/49] ext4: fix for fiemap last-block test Theodore Ts'o
2009-06-08 19:22 ` [PATCH 25/49] ext4: fix the length returned by fiemap for an unallocated extent Theodore Ts'o
2009-06-08 19:22 ` [PATCH 26/49] ext4: Convert ext4_lock_group to use sb_bgl_lock Theodore Ts'o
2009-06-08 19:22 ` [PATCH 27/49] ext4: Fix spinlock assertions on UP systems Theodore Ts'o
2009-06-08 19:22 ` [PATCH 28/49] ext4: Simplify function signature for ext4_da_get_block_write() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 29/49] ext4: Rename ext4_get_blocks_handle() to be ext4_ind_get_blocks() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 30/49] ext4: Rename ext4_get_blocks_wrap() to be ext4_get_blocks() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 31/49] ext4: Define a new set of flags for ext4_get_blocks() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 32/49] ext4: Add documentation to the ext4_*get_block* functions Theodore Ts'o
2009-06-08 19:22 ` [PATCH 33/49] ext4: Add BUG_ON debugging checks to noalloc_get_block_write() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 34/49] ext4: Merge ext4_da_get_block_write() into mpage_da_map_blocks() Theodore Ts'o
2009-06-08 19:22 ` Theodore Ts'o [this message]
2009-06-08 19:22 ` [PATCH 36/49] ext4: Add a comprehensive block validity check to ext4_get_blocks() Theodore Ts'o
2009-06-08 19:22 ` [PATCH 37/49] ext4: down i_data_sem only for read when walking tree for fiemap Theodore Ts'o
2009-06-08 19:22 ` [PATCH 38/49] ext4: Fix memory leak in ext4_fill_super() in case of a failed mount Theodore Ts'o
2009-06-08 19:22 ` [PATCH 39/49] ext3: Fix memory leak in ext3_fill_super() " Theodore Ts'o
2009-06-08 19:22 ` [PATCH 40/49] ext2: Fix memory leak in ext2_fill_super() " Theodore Ts'o
2009-06-08 19:22 ` [PATCH 41/49] ext4: remove unused function __ext4_write_dirty_metadata Theodore Ts'o
2009-06-08 19:23 ` [PATCH 42/49] ext4: Clean up calls to ext4_get_group_desc() Theodore Ts'o
2009-06-08 19:23 ` [PATCH 43/49] jbd2: Fix minor typos in comments in fs/jbd2/journal.c Theodore Ts'o
2009-06-08 19:23 ` [PATCH 44/49] ext4: super.c whitespace cleanup Theodore Ts'o
2009-06-08 19:23 ` [PATCH 45/49] ext4: Get rid of EXTEND_DISKSIZE flag of ext4_get_blocks_handle() Theodore Ts'o
2009-06-08 19:23 ` [PATCH 46/49] ext4: Change all super.c messages to print the device Theodore Ts'o
2009-06-08 19:23 ` [PATCH 47/49] ext4: Avoid leaking blocks after a block allocation failure Theodore Ts'o
2009-06-08 19:23 ` [PATCH 48/49] ext4: truncate the file properly if we fail to copy data from userspace Theodore Ts'o
2009-06-08 19:23 ` [PATCH 49/49] ext4: fix dx_map_entry to support 256k directory blocks Theodore Ts'o
2009-06-08 19:41 ` [PATCH 43/49] jbd2: Fix minor typos in comments in fs/jbd2/journal.c Alberto Bertogli
2009-06-09 4:06 ` Theodore Tso
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=1244488987-32564-36-git-send-email-tytso@mit.edu \
--to=tytso@mit.edu \
--cc=linux-kernel@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
all inboxes | Powered by JetHome®