Fix oinkmaster patch.

[ipfire-2.x.git] / src / patches / suse-2.6.27.39 / patches.suse / mnt_clone_write.patch

From: Nick Piggin <npiggin@suse.de>
Subject: fs: introduce mnt_clone_write
References: bnc#436953
Patch-upstream: no (could be submitted)

This patch speeds up lmbench lat_mmap test by about another 2% after the
first patch.

Before:
 avg = 462.286
 std = 5.46106

After:
 avg = 453.12
 std = 9.58257

(50 runs of each, stddev gives a reasonable confidence)

It does this by introducing mnt_clone_write, which avoids some heavyweight
operations of mnt_want_write if called on a vfsmount which we know already
has a write count; and mnt_want_write_file, which can call mnt_clone_write
if the file is open for write.

After these two patches, mnt_want_write and mnt_drop_write go from 7% on
the profile down to 1.3% (including mnt_clone_write).

---
 fs/file_table.c       |    2 -
 fs/inode.c            |    2 -
 fs/namespace.c        |   63 ++++++++++++++++++++++++++++++++++++++++++++++++++
 fs/open.c             |    4 +--
 fs/xattr.c            |    4 +--
 include/linux/mount.h |    5 +++
 6 files changed, 74 insertions(+), 6 deletions(-)

Index: linux-2.6.27/fs/file_table.c
===================================================================
--- linux-2.6.27.orig/fs/file_table.c
+++ linux-2.6.27/fs/file_table.c
@@ -210,7 +210,7 @@ int init_file(struct file *file, struct
 	 */
 	if ((mode & FMODE_WRITE) && !special_file(dentry->d_inode->i_mode)) {
 		file_take_write(file);
-		error = mnt_want_write(mnt);
+		error = mnt_clone_write_2(mnt);
 		WARN_ON(error);
 	}
 	return error;
Index: linux-2.6.27/fs/inode.c
===================================================================
--- linux-2.6.27.orig/fs/inode.c
+++ linux-2.6.27/fs/inode.c
@@ -1256,7 +1256,7 @@ void file_update_time(struct file *file)
 	if (IS_NOCMTIME(inode))
 		return;
 
-	err = mnt_want_write(file->f_path.mnt);
+	err = mnt_want_write_file(file->f_path.mnt, file);
 	if (err)
 		return;
 
Index: linux-2.6.27/fs/namespace.c
===================================================================
--- linux-2.6.27.orig/fs/namespace.c
+++ linux-2.6.27/fs/namespace.c
@@ -264,6 +264,69 @@ out:
 EXPORT_SYMBOL_GPL(mnt_want_write);
 
 /**
+ * mnt_clone_write - get write access to a mount
+ * @mnt: the mount on which to take a write
+ *
+ * This is effectively like mnt_want_write, except
+ * it must only be used to take an extra write reference
+ * on a mountpoint that we already know has a write reference
+ * on it. This allows some optimisation.
+ *
+ * The caller should really check __mnt_is_readonly before callint
+ * mnt_clone_write. See mnt_clone_write_2.
+ *
+ * After finished, mnt_drop_write must be called as usual to
+ * drop the reference.
+ */
+void mnt_clone_write(struct vfsmount *mnt)
+{
+	preempt_disable();
+	inc_mnt_writers(mnt);
+	preempt_enable();
+}
+EXPORT_SYMBOL_GPL(mnt_clone_write);
+
+/**
+ * mnt_clone_write_2 - get write access to a mount
+ * @mnt: the mount on which to take a write
+ *
+ * Same as mnt_clone_write, but it performs the __mnt_is_readonly
+ * check itself, and returns -error on failure. This is the preferred
+ * function. This is here to preserve kABI compatibility.
+ *
+ * After finished, mnt_drop_write must be called as usual to
+ * drop the reference.
+ */
+int mnt_clone_write_2(struct vfsmount *mnt)
+{
+	/* superblock may be r/o */
+	if (__mnt_is_readonly(mnt))
+		return -EROFS;
+	preempt_disable();
+	inc_mnt_writers(mnt);
+	preempt_enable();
+	return 0;
+}
+EXPORT_SYMBOL_GPL(mnt_clone_write_2);
+
+/**
+ * mnt_want_write_file - get write access to a file's mount
+ * @file: the file who's mount on which to take a write
+ *
+ * This is like mnt_want_write, but it takes a file and can
+ * do some optimisations if the file is open for write already
+ */
+int mnt_want_write_file(struct vfsmount *mnt, struct file *file)
+{
+	struct inode *inode = file->f_dentry->d_inode;
+	if (!(file->f_mode & FMODE_WRITE) || special_file(inode->i_mode))
+		return mnt_want_write(mnt);
+	else
+		return mnt_clone_write_2(mnt);
+}
+EXPORT_SYMBOL_GPL(mnt_want_write_file);
+
+/**
  * mnt_drop_write - give up write access to a mount
  * @mnt: the mount on which to give up write access
  *
Index: linux-2.6.27/fs/open.c
===================================================================
--- linux-2.6.27.orig/fs/open.c
+++ linux-2.6.27/fs/open.c
@@ -616,7 +616,7 @@ SYSCALL_DEFINE2(fchmod, unsigned int, fd
 
 	audit_inode(NULL, dentry);
 
-	err = mnt_want_write(file->f_path.mnt);
+	err = mnt_want_write_file(file->f_path.mnt, file);
 	if (err)
 		goto out_putf;
 	mutex_lock(&inode->i_mutex);
@@ -765,7 +765,7 @@ SYSCALL_DEFINE3(fchown, unsigned int, fd
 	if (!file)
 		goto out;
 
-	error = mnt_want_write(file->f_path.mnt);
+	error = mnt_want_write_file(file->f_path.mnt, file);
 	if (error)
 		goto out_fput;
 	dentry = file->f_path.dentry;
Index: linux-2.6.27/fs/xattr.c
===================================================================
--- linux-2.6.27.orig/fs/xattr.c
+++ linux-2.6.27/fs/xattr.c
@@ -301,7 +301,7 @@ SYSCALL_DEFINE5(fsetxattr, int, fd, cons
 		return error;
 	dentry = f->f_path.dentry;
 	audit_inode(NULL, dentry);
-	error = mnt_want_write(f->f_path.mnt);
+	error = mnt_want_write_file(f->f_path.mnt, f);
 	if (!error) {
 		error = setxattr(dentry, name, value, size, flags);
 		mnt_drop_write(f->f_path.mnt);
@@ -528,7 +528,7 @@ SYSCALL_DEFINE2(fremovexattr, int, fd, c
 		return error;
 	dentry = f->f_path.dentry;
 	audit_inode(NULL, dentry);
-	error = mnt_want_write(f->f_path.mnt);
+	error = mnt_want_write_file(f->f_path.mnt, f);
 	if (!error) {
 		error = removexattr(dentry, name);
 		mnt_drop_write(f->f_path.mnt);
Index: linux-2.6.27/include/linux/mount.h
===================================================================
--- linux-2.6.27.orig/include/linux/mount.h
+++ linux-2.6.27/include/linux/mount.h
@@ -98,7 +98,12 @@ static inline struct vfsmount *mntget(st
 	return mnt;
 }
 
+struct file;
+
 extern int mnt_want_write(struct vfsmount *mnt);
+extern int mnt_want_write_file(struct vfsmount *mnt, struct file *file);
+extern void mnt_clone_write(struct vfsmount *mnt);
+extern int mnt_clone_write_2(struct vfsmount *mnt);
 extern void mnt_drop_write(struct vfsmount *mnt);
 extern void mntput_no_expire(struct vfsmount *mnt);
 extern void mnt_pin(struct vfsmount *mnt);