[to-be-updated] ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending.patch removed from -mm tree

[to-be-updated] ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending.patch removed from -mm tree

[akpm]

The patch titled
     Subject: ocfs2: do not fallback to buffer I/O write if appending
has been removed from the -mm tree.  Its filename was
     ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending.patch

This patch was dropped because an updated version will be merged

------------------------------------------------------
From: Weiwei Wang <wangww631@huawei.com>
Subject: ocfs2: do not fallback to buffer I/O write if appending

Now we can do direct io and do not fallback to buffered IO any more in
case of append O_DIRECT write.

Signed-off-by: Weiwei Wang <wangww631@huawei.com>
Signed-off-by: Joseph Qi <joseph.qi@huawei.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 fs/ocfs2/file.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff -puN fs/ocfs2/file.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending fs/ocfs2/file.c
--- a/fs/ocfs2/file.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending
+++ a/fs/ocfs2/file.c
@@ -2116,6 +2116,9 @@ static int ocfs2_prepare_inode_for_write
 	struct dentry *dentry = file->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
 	loff_t saved_pos = 0, end;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	int full_coherency = !(osb->s_mount_opt &
+		OCFS2_MOUNT_COHERENCY_BUFFERED);
 
 	/*
 	 * We start with a read level meta lock and only jump to an ex
@@ -2204,7 +2207,7 @@ static int ocfs2_prepare_inode_for_write
 		 * one node could wind up truncating another
 		 * nodes writes.
 		 */
-		if (end > i_size_read(inode)) {
+		if (end > i_size_read(inode) && !full_coherency) {
 			*direct_io = 0;
 			break;
 		}
_

Patches currently in -mm which might be from wangww631@huawei.com are

ocfs2-add-a-mount-option-journal_async_commit-on-ocfs2-filesystem.patch
ocfs2-do-not-fallback-to-buffer-i-o-write-if-fill-holes.patch
ocfs2-fix-leftover-orphan-entry-caused-by-append-o_direct-write-crash.patch
ocfs2-implement-ocfs2_direct_io_write-fix.patch

[to-be-updated] ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending.patch removed from -mm tree

[akpm]

The patch titled
     Subject: ocfs2: do not fallback to buffer I/O write if appending
has been removed from the -mm tree.  Its filename was
     ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending.patch

This patch was dropped because an updated version will be merged

------------------------------------------------------
From: WeiWei Wang <wangww631@huawei.com>
Subject: ocfs2: do not fallback to buffer I/O write if appending

Now we can do direct io and do not fallback to buffered IO any more in
case of append O_DIRECT write.

Signed-off-by: Weiwei Wang <wangww631@huawei.com>
Reviewed-by: Joseph Qi <joseph.qi@huawei.com>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 fs/ocfs2/file.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff -puN fs/ocfs2/file.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending fs/ocfs2/file.c
--- a/fs/ocfs2/file.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending
+++ a/fs/ocfs2/file.c
@@ -2109,6 +2109,9 @@ static int ocfs2_prepare_inode_for_write
 	struct dentry *dentry = file->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
 	loff_t saved_pos = 0, end;
+	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+	int full_coherency = !(osb->s_mount_opt &
+			OCFS2_MOUNT_COHERENCY_BUFFERED);
 
 	/*
 	 * We start with a read level meta lock and only jump to an ex
@@ -2197,7 +2200,7 @@ static int ocfs2_prepare_inode_for_write
 		 * one node could wind up truncating another
 		 * nodes writes.
 		 */
-		if (end > i_size_read(inode)) {
+		if (end > i_size_read(inode) && !full_coherency) {
 			*direct_io = 0;
 			break;
 		}
_

Patches currently in -mm which might be from wangww631@huawei.com are

ocfs2-do-not-fallback-to-buffer-i-o-write-if-fill-holes.patch
ocfs2-do-not-fallback-to-buffer-i-o-write-if-fill-holes-checkpatch-fixes.patch

[to-be-updated] ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending.patch removed from -mm tree

[akpm]

The patch titled
     Subject: ocfs2: do not fall back to buffer I/O write if appending
has been removed from the -mm tree.  Its filename was
     ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending.patch

This patch was dropped because an updated version will be merged

------------------------------------------------------
From: WeiWei Wang <wangww631@huawei.com>
Subject: ocfs2: do not fall back to buffer I/O write if appending

After enabling block allocation in direct-io write, we don't fall back to
buffer I/O write in append write situation any more.  If the O_DIRECT flag
is taken, we use direct-io.

Signed-off-by: Weiwei Wang <wangww631@huawei.com>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 fs/ocfs2/aops.c |    4 ----
 fs/ocfs2/file.c |   11 -----------
 2 files changed, 15 deletions(-)

diff -puN fs/ocfs2/aops.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending fs/ocfs2/aops.c
--- a/fs/ocfs2/aops.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending
+++ a/fs/ocfs2/aops.c
@@ -663,10 +663,6 @@ static ssize_t ocfs2_direct_IO(int rw,
 	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL)
 		return 0;
 
-	/* Fallback to buffered I/O if we are appending. */
-	if (i_size_read(inode) <= offset)
-		return 0;
-
 	if (rw == WRITE) {
 		loff_t final_size = offset + count;
 		/*
diff -puN fs/ocfs2/file.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending fs/ocfs2/file.c
--- a/fs/ocfs2/file.c~ocfs2-do-not-fallback-to-buffer-i-o-write-if-appending
+++ a/fs/ocfs2/file.c
@@ -2192,17 +2192,6 @@ static int ocfs2_prepare_inode_for_write
 		}
 
 		/*
-		 * Allowing concurrent direct writes means
-		 * i_size changes wouldn't be synchronized, so
-		 * one node could wind up truncating another
-		 * nodes writes.
-		 */
-		if (end > i_size_read(inode)) {
-			*direct_io = 0;
-			break;
-		}