From 44396f4b5cb8566f7118aec55eeac99be7ad94cb Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Tue, 31 May 2011 11:58:49 -0400
Subject: fs: add a DCACHE_NEED_LOOKUP flag for d_flags

Btrfs (and I'd venture most other fs's) stores its indexes in nice disk order
for readdir, but unfortunately in the case of anything that stats the files in
order that readdir spits back (like oh say ls) that means we still have to do
the normal lookup of the file, which means looking up our other index and then
looking up the inode.  What I want is a way to create dummy dentries when we
find them in readdir so that when ls or anything else subsequently does a
stat(), we already have the location information in the dentry and can go
straight to the inode itself.  The lookup stuff just assumes that if it finds a
dentry it is done, it doesn't perform a lookup.  So add a DCACHE_NEED_LOOKUP
flag so that the lookup code knows it still needs to run i_op->lookup() on the
parent to get the inode for the dentry.  I have tested this with btrfs and I
went from something that looks like this

http://people.redhat.com/jwhiter/ls-noreada.png

To this

http://people.redhat.com/jwhiter/ls-good.png

Thats a savings of 1300 seconds, or 22 minutes.  That is a significant savings.
Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c | 34 ++++++++++++++++++++++++++++++++--
 fs/namei.c  | 49 +++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 81 insertions(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/dcache.c b/fs/dcache.c
index 6e4ea6d87774..d3902139b533 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -343,6 +343,24 @@ void d_drop(struct dentry *dentry)
 }
 EXPORT_SYMBOL(d_drop);
 
+/*
+ * d_clear_need_lookup - drop a dentry from cache and clear the need lookup flag
+ * @dentry: dentry to drop
+ *
+ * This is called when we do a lookup on a placeholder dentry that needed to be
+ * looked up.  The dentry should have been hashed in order for it to be found by
+ * the lookup code, but now needs to be unhashed while we do the actual lookup
+ * and clear the DCACHE_NEED_LOOKUP flag.
+ */
+void d_clear_need_lookup(struct dentry *dentry)
+{
+	spin_lock(&dentry->d_lock);
+	__d_drop(dentry);
+	dentry->d_flags &= ~DCACHE_NEED_LOOKUP;
+	spin_unlock(&dentry->d_lock);
+}
+EXPORT_SYMBOL(d_clear_need_lookup);
+
 /*
  * Finish off a dentry we've decided to kill.
  * dentry->d_lock must be held, returns with it unlocked.
@@ -432,8 +450,13 @@ repeat:
  	if (d_unhashed(dentry))
 		goto kill_it;
 
-	/* Otherwise leave it cached and ensure it's on the LRU */
-	dentry->d_flags |= DCACHE_REFERENCED;
+	/*
+	 * If this dentry needs lookup, don't set the referenced flag so that it
+	 * is more likely to be cleaned up by the dcache shrinker in case of
+	 * memory pressure.
+	 */
+	if (!d_need_lookup(dentry))
+		dentry->d_flags |= DCACHE_REFERENCED;
 	dentry_lru_add(dentry);
 
 	dentry->d_count--;
@@ -1707,6 +1730,13 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
 		return found;
 	}
 
+	/*
+	 * We are going to instantiate this dentry, unhash it and clear the
+	 * lookup flag so we can do that.
+	 */
+	if (unlikely(d_need_lookup(found)))
+		d_clear_need_lookup(found);
+
 	/*
 	 * Negative dentry: instantiate it unless the inode is a directory and
 	 * already has a dentry.
diff --git a/fs/namei.c b/fs/namei.c
index 14ab8d3f2f0c..5ba42c453e31 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1133,6 +1133,30 @@ static struct dentry *d_alloc_and_lookup(struct dentry *parent,
 	return dentry;
 }
 
+/*
+ * We already have a dentry, but require a lookup to be performed on the parent
+ * directory to fill in d_inode. Returns the new dentry, or ERR_PTR on error.
+ * parent->d_inode->i_mutex must be held. d_lookup must have verified that no
+ * child exists while under i_mutex.
+ */
+static struct dentry *d_inode_lookup(struct dentry *parent, struct dentry *dentry,
+				     struct nameidata *nd)
+{
+	struct inode *inode = parent->d_inode;
+	struct dentry *old;
+
+	/* Don't create child dentry for a dead directory. */
+	if (unlikely(IS_DEADDIR(inode)))
+		return ERR_PTR(-ENOENT);
+
+	old = inode->i_op->lookup(inode, dentry, nd);
+	if (unlikely(old)) {
+		dput(dentry);
+		dentry = old;
+	}
+	return dentry;
+}
+
 /*
  *  It's more convoluted than I'd like it to be, but... it's still fairly
  *  small and for now I'd prefer to have fast path as straight as possible.
@@ -1172,6 +1196,8 @@ static int do_lookup(struct nameidata *nd, struct qstr *name,
 				goto unlazy;
 			}
 		}
+		if (unlikely(d_need_lookup(dentry)))
+			goto unlazy;
 		path->mnt = mnt;
 		path->dentry = dentry;
 		if (unlikely(!__follow_mount_rcu(nd, path, inode)))
@@ -1186,6 +1212,10 @@ unlazy:
 		dentry = __d_lookup(parent, name);
 	}
 
+	if (dentry && unlikely(d_need_lookup(dentry))) {
+		dput(dentry);
+		dentry = NULL;
+	}
 retry:
 	if (unlikely(!dentry)) {
 		struct inode *dir = parent->d_inode;
@@ -1202,6 +1232,15 @@ retry:
 			/* known good */
 			need_reval = 0;
 			status = 1;
+		} else if (unlikely(d_need_lookup(dentry))) {
+			dentry = d_inode_lookup(parent, dentry, nd);
+			if (IS_ERR(dentry)) {
+				mutex_unlock(&dir->i_mutex);
+				return PTR_ERR(dentry);
+			}
+			/* known good */
+			need_reval = 0;
+			status = 1;
 		}
 		mutex_unlock(&dir->i_mutex);
 	}
@@ -1683,6 +1722,16 @@ static struct dentry *__lookup_hash(struct qstr *name,
 	 */
 	dentry = d_lookup(base, name);
 
+	if (dentry && d_need_lookup(dentry)) {
+		/*
+		 * __lookup_hash is called with the parent dir's i_mutex already
+		 * held, so we are good to go here.
+		 */
+		dentry = d_inode_lookup(base, dentry, nd);
+		if (IS_ERR(dentry))
+			return dentry;
+	}
+
 	if (dentry && (dentry->d_flags & DCACHE_OP_REVALIDATE))
 		dentry = do_revalidate(dentry, nd);
 
-- 
cgit v1.2.3


From 43e15cdbefea4ce6d68113de98d4f61c0cf45687 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 3 Jun 2011 20:16:57 -0400
Subject: new helper: iterate_supers_type()

Call the given function for all superblocks of given type.  Function
gets a superblock (with s_umount locked shared) and (void *) argument
supplied by caller of iterator.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/super.c | 36 ++++++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)

(limited to 'fs')

diff --git a/fs/super.c b/fs/super.c
index ab3d672db0de..444da9579068 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -451,6 +451,42 @@ void iterate_supers(void (*f)(struct super_block *, void *), void *arg)
 	spin_unlock(&sb_lock);
 }
 
+/**
+ *	iterate_supers_type - call function for superblocks of given type
+ *	@type: fs type
+ *	@f: function to call
+ *	@arg: argument to pass to it
+ *
+ *	Scans the superblock list and calls given function, passing it
+ *	locked superblock and given argument.
+ */
+void iterate_supers_type(struct file_system_type *type,
+	void (*f)(struct super_block *, void *), void *arg)
+{
+	struct super_block *sb, *p = NULL;
+
+	spin_lock(&sb_lock);
+	list_for_each_entry(sb, &type->fs_supers, s_instances) {
+		sb->s_count++;
+		spin_unlock(&sb_lock);
+
+		down_read(&sb->s_umount);
+		if (sb->s_root)
+			f(sb, arg);
+		up_read(&sb->s_umount);
+
+		spin_lock(&sb_lock);
+		if (p)
+			__put_super(p);
+		p = sb;
+	}
+	if (p)
+		__put_super(p);
+	spin_unlock(&sb_lock);
+}
+
+EXPORT_SYMBOL(iterate_supers_type);
+
 /**
  *	get_super - get the superblock of a device
  *	@bdev: device to get the superblock for
-- 
cgit v1.2.3


From 4cf27141cbe0239f48ec6f0b37bad347d51d1785 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 19 Jun 2011 01:50:08 -0400
Subject: make exec_permission(dir) really equivalent to inode_permission(dir,
 MAY_EXEC)

capability overrides apply only to the default case; if fs has ->permission()
that does _not_ call generic_permission(), we have no business doing them.
Moreover, if it has ->permission() that does call generic_permission(), we
have no need to recheck capabilities.

Besides, the capability overrides should apply only if we got EACCES from
acl_permission_check(); any other value (-EIO, etc.) should be returned
to caller, capabilities or not capabilities.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 5ba42c453e31..7c8a93042e63 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -584,19 +584,19 @@ static inline int exec_permission(struct inode *inode, unsigned int flags)
 
 	if (inode->i_op->permission) {
 		ret = inode->i_op->permission(inode, MAY_EXEC, flags);
+		if (likely(!ret))
+			goto ok;
 	} else {
 		ret = acl_permission_check(inode, MAY_EXEC, flags,
 				inode->i_op->check_acl);
+		if (likely(!ret))
+			goto ok;
+		if (ret != -EACCES)
+			return ret;
+		if (ns_capable(ns, CAP_DAC_OVERRIDE) ||
+				ns_capable(ns, CAP_DAC_READ_SEARCH))
+			goto ok;
 	}
-	if (likely(!ret))
-		goto ok;
-	if (ret == -ECHILD)
-		return ret;
-
-	if (ns_capable(ns, CAP_DAC_OVERRIDE) ||
-			ns_capable(ns, CAP_DAC_READ_SEARCH))
-		goto ok;
-
 	return ret;
 ok:
 	return security_inode_exec_permission(inode, flags);
-- 
cgit v1.2.3


From 6f2861097467852f2271c2b40f9c3d1d01757048 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 19 Jun 2011 11:49:08 -0400
Subject: switch udf_ioctl() to inode_permission()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/udf/file.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/udf/file.c b/fs/udf/file.c
index 2a346bb1d9f5..d8ffa7cc661d 100644
--- a/fs/udf/file.c
+++ b/fs/udf/file.c
@@ -150,7 +150,7 @@ long udf_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	long old_block, new_block;
 	int result = -EINVAL;
 
-	if (file_permission(filp, MAY_READ) != 0) {
+	if (inode_permission(inode, MAY_READ) != 0) {
 		udf_debug("no permission to access inode %lu\n", inode->i_ino);
 		result = -EPERM;
 		goto out;
-- 
cgit v1.2.3


From 78f32a9b479e9b9f1ce2bf620a7602c1cdbc4c8e Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 19 Jun 2011 11:54:42 -0400
Subject: switch path_init() to exec_permission()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 7c8a93042e63..cf2554635a1c 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1554,7 +1554,7 @@ static int path_init(int dfd, const char *name, unsigned int flags,
 			if (!S_ISDIR(dentry->d_inode->i_mode))
 				goto fput_fail;
 
-			retval = file_permission(file, MAY_EXEC);
+			retval = exec_permission(dentry->d_inode, 0);
 			if (retval)
 				goto fput_fail;
 		}
-- 
cgit v1.2.3


From 1b5d783c94c328d406e801566f161adcfb018dda Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 19 Jun 2011 12:49:47 -0400
Subject: consolidate BINPRM_FLAGS_ENFORCE_NONDUMP handling

new helper: would_dump(bprm, file).  Checks if we are allowed to
read the file and if we are not - sets ENFORCE_NODUMP.  Exported,
used in places that previously open-coded the same logics.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/binfmt_elf.c       |  3 +--
 fs/binfmt_elf_fdpic.c |  3 +--
 fs/binfmt_misc.c      |  3 +--
 fs/exec.c             | 14 +++++++++++---
 4 files changed, 14 insertions(+), 9 deletions(-)

(limited to 'fs')

diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 303983fabfd6..dd0fdfc56d38 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -668,8 +668,7 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 			 * mm->dumpable = 0 regardless of the interpreter's
 			 * permissions.
 			 */
-			if (file_permission(interpreter, MAY_READ) < 0)
-				bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+			would_dump(bprm, interpreter);
 
 			retval = kernel_read(interpreter, 0, bprm->buf,
 					     BINPRM_BUF_SIZE);
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index 2bc5dc644b4c..30745f459faf 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -245,8 +245,7 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm,
 			 * mm->dumpable = 0 regardless of the interpreter's
 			 * permissions.
 			 */
-			if (file_permission(interpreter, MAY_READ) < 0)
-				bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+			would_dump(bprm, interpreter);
 
 			retval = kernel_read(interpreter, 0, bprm->buf,
 					     BINPRM_BUF_SIZE);
diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c
index 1befe2ec8186..ba1a1ae4a18a 100644
--- a/fs/binfmt_misc.c
+++ b/fs/binfmt_misc.c
@@ -149,8 +149,7 @@ static int load_misc_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 
 		/* if the binary is not readable than enforce mm->dumpable=0
 		   regardless of the interpreter's permissions */
-		if (file_permission(bprm->file, MAY_READ))
-			bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+		would_dump(bprm, bprm->file);
 
 		allow_write_access(bprm->file);
 		bprm->file = NULL;
diff --git a/fs/exec.c b/fs/exec.c
index 6075a1e727ae..f9f12ad299af 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -1105,6 +1105,13 @@ out:
 }
 EXPORT_SYMBOL(flush_old_exec);
 
+void would_dump(struct linux_binprm *bprm, struct file *file)
+{
+	if (inode_permission(file->f_path.dentry->d_inode, MAY_READ) < 0)
+		bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
+}
+EXPORT_SYMBOL(would_dump);
+
 void setup_new_exec(struct linux_binprm * bprm)
 {
 	int i, ch;
@@ -1144,9 +1151,10 @@ void setup_new_exec(struct linux_binprm * bprm)
 	if (bprm->cred->uid != current_euid() ||
 	    bprm->cred->gid != current_egid()) {
 		current->pdeath_signal = 0;
-	} else if (file_permission(bprm->file, MAY_READ) ||
-		   bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP) {
-		set_dumpable(current->mm, suid_dumpable);
+	} else {
+		would_dump(bprm, bprm->file);
+		if (bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP)
+			set_dumpable(current->mm, suid_dumpable);
 	}
 
 	/*
-- 
cgit v1.2.3


From 3bfa784a6539f91a27d7ffdd408efdb638e3bebd Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 19 Jun 2011 12:55:10 -0400
Subject: kill file_permission() completely

convert the last remaining caller to inode_permission()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 18 ------------------
 1 file changed, 18 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index cf2554635a1c..4ad2b781a65c 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -303,23 +303,6 @@ int inode_permission(struct inode *inode, int mask)
 	return security_inode_permission(inode, mask);
 }
 
-/**
- * file_permission  -  check for additional access rights to a given file
- * @file:	file to check access rights for
- * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
- *
- * Used to check for read/write/execute permissions on an already opened
- * file.
- *
- * Note:
- *	Do not use this function in new code.  All access checks should
- *	be done using inode_permission().
- */
-int file_permission(struct file *file, int mask)
-{
-	return inode_permission(file->f_path.dentry->d_inode, mask);
-}
-
 /*
  * get_write_access() gets write permission for a file.
  * put_write_access() releases this write permission.
@@ -3405,7 +3388,6 @@ EXPORT_SYMBOL(kern_path_parent);
 EXPORT_SYMBOL(kern_path);
 EXPORT_SYMBOL(vfs_path_lookup);
 EXPORT_SYMBOL(inode_permission);
-EXPORT_SYMBOL(file_permission);
 EXPORT_SYMBOL(unlock_rename);
 EXPORT_SYMBOL(vfs_create);
 EXPORT_SYMBOL(vfs_follow_link);
-- 
cgit v1.2.3


From f4d6ff89d8e54b68a4322388d26d518d6133fa4e Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 19 Jun 2011 13:14:21 -0400
Subject: move exec_permission() up to the rest of permission-related functions

... and convert the comment before it into linuxdoc form.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 72 +++++++++++++++++++++++++++++++++-----------------------------
 1 file changed, 38 insertions(+), 34 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 4ad2b781a65c..e04f15ae4b07 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -303,6 +303,44 @@ int inode_permission(struct inode *inode, int mask)
 	return security_inode_permission(inode, mask);
 }
 
+/**
+ * exec_permission  -  check for right to do lookups in a given directory
+ * @inode:	inode to check permission on
+ * @flags:	IPERM_FLAG_ flags.
+ *
+ * Short-cut version of inode_permission(), for calling on directories
+ * during pathname resolution.  Combines parts of inode_permission()
+ * and generic_permission(), and tests ONLY for MAY_EXEC permission.
+ *
+ * If appropriate, check DAC only.  If not appropriate, or
+ * short-cut DAC fails, then call ->permission() to do more
+ * complete permission check.
+ */
+static inline int exec_permission(struct inode *inode, unsigned int flags)
+{
+	int ret;
+	struct user_namespace *ns = inode_userns(inode);
+
+	if (inode->i_op->permission) {
+		ret = inode->i_op->permission(inode, MAY_EXEC, flags);
+		if (likely(!ret))
+			goto ok;
+	} else {
+		ret = acl_permission_check(inode, MAY_EXEC, flags,
+				inode->i_op->check_acl);
+		if (likely(!ret))
+			goto ok;
+		if (ret != -EACCES)
+			return ret;
+		if (ns_capable(ns, CAP_DAC_OVERRIDE) ||
+				ns_capable(ns, CAP_DAC_READ_SEARCH))
+			goto ok;
+	}
+	return ret;
+ok:
+	return security_inode_exec_permission(inode, flags);
+}
+
 /*
  * get_write_access() gets write permission for a file.
  * put_write_access() releases this write permission.
@@ -551,40 +589,6 @@ static int complete_walk(struct nameidata *nd)
 	return status;
 }
 
-/*
- * Short-cut version of permission(), for calling on directories
- * during pathname resolution.  Combines parts of permission()
- * and generic_permission(), and tests ONLY for MAY_EXEC permission.
- *
- * If appropriate, check DAC only.  If not appropriate, or
- * short-cut DAC fails, then call ->permission() to do more
- * complete permission check.
- */
-static inline int exec_permission(struct inode *inode, unsigned int flags)
-{
-	int ret;
-	struct user_namespace *ns = inode_userns(inode);
-
-	if (inode->i_op->permission) {
-		ret = inode->i_op->permission(inode, MAY_EXEC, flags);
-		if (likely(!ret))
-			goto ok;
-	} else {
-		ret = acl_permission_check(inode, MAY_EXEC, flags,
-				inode->i_op->check_acl);
-		if (likely(!ret))
-			goto ok;
-		if (ret != -EACCES)
-			return ret;
-		if (ns_capable(ns, CAP_DAC_OVERRIDE) ||
-				ns_capable(ns, CAP_DAC_READ_SEARCH))
-			goto ok;
-	}
-	return ret;
-ok:
-	return security_inode_exec_permission(inode, flags);
-}
-
 static __always_inline void set_root(struct nameidata *nd)
 {
 	if (!nd->root.mnt)
-- 
cgit v1.2.3


From 07b8ce1ee87d291ff564c02cf878fae973317a52 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 10:52:57 -0400
Subject: lockless get_write_access/deny_write_access

new helpers: atomic_inc_unless_negative()/atomic_dec_unless_positive()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 46 ----------------------------------------------
 1 file changed, 46 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index e04f15ae4b07..d286cbc3f3e5 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -341,52 +341,6 @@ ok:
 	return security_inode_exec_permission(inode, flags);
 }
 
-/*
- * get_write_access() gets write permission for a file.
- * put_write_access() releases this write permission.
- * This is used for regular files.
- * We cannot support write (and maybe mmap read-write shared) accesses and
- * MAP_DENYWRITE mmappings simultaneously. The i_writecount field of an inode
- * can have the following values:
- * 0: no writers, no VM_DENYWRITE mappings
- * < 0: (-i_writecount) vm_area_structs with VM_DENYWRITE set exist
- * > 0: (i_writecount) users are writing to the file.
- *
- * Normally we operate on that counter with atomic_{inc,dec} and it's safe
- * except for the cases where we don't hold i_writecount yet. Then we need to
- * use {get,deny}_write_access() - these functions check the sign and refuse
- * to do the change if sign is wrong. Exclusion between them is provided by
- * the inode->i_lock spinlock.
- */
-
-int get_write_access(struct inode * inode)
-{
-	spin_lock(&inode->i_lock);
-	if (atomic_read(&inode->i_writecount) < 0) {
-		spin_unlock(&inode->i_lock);
-		return -ETXTBSY;
-	}
-	atomic_inc(&inode->i_writecount);
-	spin_unlock(&inode->i_lock);
-
-	return 0;
-}
-
-int deny_write_access(struct file * file)
-{
-	struct inode *inode = file->f_path.dentry->d_inode;
-
-	spin_lock(&inode->i_lock);
-	if (atomic_read(&inode->i_writecount) > 0) {
-		spin_unlock(&inode->i_lock);
-		return -ETXTBSY;
-	}
-	atomic_dec(&inode->i_writecount);
-	spin_unlock(&inode->i_lock);
-
-	return 0;
-}
-
 /**
  * path_get - get a reference to a path
  * @path: path to get the reference to
-- 
cgit v1.2.3


From 178ea73521d64ba41d7aa5488fb9f549c6d4507d Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 11:31:30 -0400
Subject: kill check_acl callback of generic_permission()

its value depends only on inode and does not change; we might as
well store it in ->i_op->check_acl and be done with that.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/afs/security.c       |  2 +-
 fs/btrfs/inode.c        |  7 ++++++-
 fs/ceph/inode.c         |  2 +-
 fs/cifs/cifsfs.c        |  2 +-
 fs/fuse/dir.c           |  4 ++--
 fs/gfs2/inode.c         |  5 ++++-
 fs/hostfs/hostfs_kern.c |  2 +-
 fs/hpfs/namei.c         |  2 +-
 fs/namei.c              | 17 +++++++----------
 fs/nfs/dir.c            |  2 +-
 fs/nilfs2/inode.c       |  2 +-
 fs/ocfs2/file.c         |  4 +++-
 fs/ocfs2/namei.c        |  1 +
 fs/proc/base.c          |  2 +-
 fs/reiserfs/file.c      |  1 +
 fs/reiserfs/namei.c     |  4 +++-
 fs/reiserfs/xattr.c     | 18 ++++++++----------
 fs/sysfs/inode.c        |  2 +-
 18 files changed, 44 insertions(+), 35 deletions(-)

(limited to 'fs')

diff --git a/fs/afs/security.c b/fs/afs/security.c
index f44b9d355377..745ee654165f 100644
--- a/fs/afs/security.c
+++ b/fs/afs/security.c
@@ -350,7 +350,7 @@ int afs_permission(struct inode *inode, int mask, unsigned int flags)
 	}
 
 	key_put(key);
-	ret = generic_permission(inode, mask, flags, NULL);
+	ret = generic_permission(inode, mask, flags);
 	_leave(" = %d", ret);
 	return ret;
 
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 3601f0aebddf..f0bd87371566 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -7339,7 +7339,7 @@ static int btrfs_permission(struct inode *inode, int mask, unsigned int flags)
 		return -EROFS;
 	if ((BTRFS_I(inode)->flags & BTRFS_INODE_READONLY) && (mask & MAY_WRITE))
 		return -EACCES;
-	return generic_permission(inode, mask, flags, btrfs_check_acl);
+	return generic_permission(inode, mask, flags);
 }
 
 static const struct inode_operations btrfs_dir_inode_operations = {
@@ -7359,10 +7359,12 @@ static const struct inode_operations btrfs_dir_inode_operations = {
 	.listxattr	= btrfs_listxattr,
 	.removexattr	= btrfs_removexattr,
 	.permission	= btrfs_permission,
+	.check_acl	= btrfs_check_acl,
 };
 static const struct inode_operations btrfs_dir_ro_inode_operations = {
 	.lookup		= btrfs_lookup,
 	.permission	= btrfs_permission,
+	.check_acl	= btrfs_check_acl,
 };
 
 static const struct file_operations btrfs_dir_file_operations = {
@@ -7431,6 +7433,7 @@ static const struct inode_operations btrfs_file_inode_operations = {
 	.removexattr	= btrfs_removexattr,
 	.permission	= btrfs_permission,
 	.fiemap		= btrfs_fiemap,
+	.check_acl	= btrfs_check_acl,
 };
 static const struct inode_operations btrfs_special_inode_operations = {
 	.getattr	= btrfs_getattr,
@@ -7440,6 +7443,7 @@ static const struct inode_operations btrfs_special_inode_operations = {
 	.getxattr	= btrfs_getxattr,
 	.listxattr	= btrfs_listxattr,
 	.removexattr	= btrfs_removexattr,
+	.check_acl	= btrfs_check_acl,
 };
 static const struct inode_operations btrfs_symlink_inode_operations = {
 	.readlink	= generic_readlink,
@@ -7451,6 +7455,7 @@ static const struct inode_operations btrfs_symlink_inode_operations = {
 	.getxattr	= btrfs_getxattr,
 	.listxattr	= btrfs_listxattr,
 	.removexattr	= btrfs_removexattr,
+	.check_acl	= btrfs_check_acl,
 };
 
 const struct dentry_operations btrfs_dentry_operations = {
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index d8858e96ab18..beb5d55d6fd2 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -1805,7 +1805,7 @@ int ceph_permission(struct inode *inode, int mask, unsigned int flags)
 	err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED);
 
 	if (!err)
-		err = generic_permission(inode, mask, flags, NULL);
+		err = generic_permission(inode, mask, flags);
 	return err;
 }
 
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index bc4b12ca537b..b79804fa410f 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -239,7 +239,7 @@ static int cifs_permission(struct inode *inode, int mask, unsigned int flags)
 		on the client (above and beyond ACL on servers) for
 		servers which do not support setting and viewing mode bits,
 		so allowing client to check permissions is useful */
-		return generic_permission(inode, mask, flags, NULL);
+		return generic_permission(inode, mask, flags);
 }
 
 static struct kmem_cache *cifs_inode_cachep;
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index d50160714595..0a2fcd860ad6 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -1018,7 +1018,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 	}
 
 	if (fc->flags & FUSE_DEFAULT_PERMISSIONS) {
-		err = generic_permission(inode, mask, flags, NULL);
+		err = generic_permission(inode, mask, flags);
 
 		/* If permission is denied, try to refresh file
 		   attributes.  This is also needed, because the root
@@ -1027,7 +1027,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 			err = fuse_perm_getattr(inode, flags);
 			if (!err)
 				err = generic_permission(inode, mask,
-							flags, NULL);
+							flags);
 		}
 
 		/* Note: the opposite of the above test does not
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 03e0c529063e..d5f0f4ea25dc 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -1564,7 +1564,7 @@ int gfs2_permission(struct inode *inode, int mask, unsigned int flags)
 	if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode))
 		error = -EACCES;
 	else
-		error = generic_permission(inode, mask, flags, gfs2_check_acl);
+		error = generic_permission(inode, mask, flags);
 	if (unlock)
 		gfs2_glock_dq_uninit(&i_gh);
 
@@ -1854,6 +1854,7 @@ const struct inode_operations gfs2_file_iops = {
 	.listxattr = gfs2_listxattr,
 	.removexattr = gfs2_removexattr,
 	.fiemap = gfs2_fiemap,
+	.check_acl = gfs2_check_acl,
 };
 
 const struct inode_operations gfs2_dir_iops = {
@@ -1874,6 +1875,7 @@ const struct inode_operations gfs2_dir_iops = {
 	.listxattr = gfs2_listxattr,
 	.removexattr = gfs2_removexattr,
 	.fiemap = gfs2_fiemap,
+	.check_acl = gfs2_check_acl,
 };
 
 const struct inode_operations gfs2_symlink_iops = {
@@ -1888,5 +1890,6 @@ const struct inode_operations gfs2_symlink_iops = {
 	.listxattr = gfs2_listxattr,
 	.removexattr = gfs2_removexattr,
 	.fiemap = gfs2_fiemap,
+	.check_acl = gfs2_check_acl,
 };
 
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index 2638c834ed28..a98d0d1aef65 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -770,7 +770,7 @@ int hostfs_permission(struct inode *ino, int desired, unsigned int flags)
 		err = access_file(name, r, w, x);
 	__putname(name);
 	if (!err)
-		err = generic_permission(ino, desired, flags, NULL);
+		err = generic_permission(ino, desired, flags);
 	return err;
 }
 
diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c
index acf95dab2aac..bd2ce7dd8df3 100644
--- a/fs/hpfs/namei.c
+++ b/fs/hpfs/namei.c
@@ -398,7 +398,7 @@ again:
 			hpfs_unlock(dir->i_sb);
 			return -ENOSPC;
 		}
-		if (generic_permission(inode, MAY_WRITE, 0, NULL) ||
+		if (generic_permission(inode, MAY_WRITE, 0) ||
 		    !S_ISREG(inode->i_mode) ||
 		    get_write_access(inode)) {
 			d_rehash(dentry);
diff --git a/fs/namei.c b/fs/namei.c
index d286cbc3f3e5..c5af0f37e67d 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -176,9 +176,9 @@ EXPORT_SYMBOL(putname);
 /*
  * This does basic POSIX ACL permission checking
  */
-static int acl_permission_check(struct inode *inode, int mask, unsigned int flags,
-		int (*check_acl)(struct inode *inode, int mask, unsigned int flags))
+static int acl_permission_check(struct inode *inode, int mask, unsigned int flags)
 {
+	int (*check_acl)(struct inode *inode, int mask, unsigned int flags);
 	unsigned int mode = inode->i_mode;
 
 	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
@@ -189,6 +189,7 @@ static int acl_permission_check(struct inode *inode, int mask, unsigned int flag
 	if (current_fsuid() == inode->i_uid)
 		mode >>= 6;
 	else {
+		check_acl = inode->i_op->check_acl;
 		if (IS_POSIXACL(inode) && (mode & S_IRWXG) && check_acl) {
 			int error = check_acl(inode, mask, flags);
 			if (error != -EAGAIN)
@@ -212,7 +213,6 @@ other_perms:
  * generic_permission -  check for access rights on a Posix-like filesystem
  * @inode:	inode to check access rights for
  * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
- * @check_acl:	optional callback to check for Posix ACLs
  * @flags:	IPERM_FLAG_ flags.
  *
  * Used to check for read/write/execute permissions on a file.
@@ -224,15 +224,14 @@ other_perms:
  * request cannot be satisfied (eg. requires blocking or too much complexity).
  * It would then be called again in ref-walk mode.
  */
-int generic_permission(struct inode *inode, int mask, unsigned int flags,
-	int (*check_acl)(struct inode *inode, int mask, unsigned int flags))
+int generic_permission(struct inode *inode, int mask, unsigned int flags)
 {
 	int ret;
 
 	/*
 	 * Do the basic POSIX ACL permission checks.
 	 */
-	ret = acl_permission_check(inode, mask, flags, check_acl);
+	ret = acl_permission_check(inode, mask, flags);
 	if (ret != -EACCES)
 		return ret;
 
@@ -290,8 +289,7 @@ int inode_permission(struct inode *inode, int mask)
 	if (inode->i_op->permission)
 		retval = inode->i_op->permission(inode, mask, 0);
 	else
-		retval = generic_permission(inode, mask, 0,
-				inode->i_op->check_acl);
+		retval = generic_permission(inode, mask, 0);
 
 	if (retval)
 		return retval;
@@ -326,8 +324,7 @@ static inline int exec_permission(struct inode *inode, unsigned int flags)
 		if (likely(!ret))
 			goto ok;
 	} else {
-		ret = acl_permission_check(inode, MAY_EXEC, flags,
-				inode->i_op->check_acl);
+		ret = acl_permission_check(inode, MAY_EXEC, flags);
 		if (likely(!ret))
 			goto ok;
 		if (ret != -EACCES)
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index ededdbd0db38..0485dca34fb1 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -2328,7 +2328,7 @@ out:
 out_notsup:
 	res = nfs_revalidate_inode(NFS_SERVER(inode), inode);
 	if (res == 0)
-		res = generic_permission(inode, mask, flags, NULL);
+		res = generic_permission(inode, mask, flags);
 	goto out;
 }
 
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index b9b45fc2903e..650aa7755003 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -806,7 +806,7 @@ int nilfs_permission(struct inode *inode, int mask, unsigned int flags)
 	    root->cno != NILFS_CPTREE_CURRENT_CNO)
 		return -EROFS; /* snapshot is not writable */
 
-	return generic_permission(inode, mask, flags, NULL);
+	return generic_permission(inode, mask, flags);
 }
 
 int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh)
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index b1e35a392ca5..d058cb7e12d4 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1293,7 +1293,7 @@ int ocfs2_permission(struct inode *inode, int mask, unsigned int flags)
 		goto out;
 	}
 
-	ret = generic_permission(inode, mask, flags, ocfs2_check_acl);
+	ret = generic_permission(inode, mask, flags);
 
 	ocfs2_inode_unlock(inode, 0);
 out:
@@ -2593,12 +2593,14 @@ const struct inode_operations ocfs2_file_iops = {
 	.listxattr	= ocfs2_listxattr,
 	.removexattr	= generic_removexattr,
 	.fiemap		= ocfs2_fiemap,
+	.check_acl	= ocfs2_check_acl,
 };
 
 const struct inode_operations ocfs2_special_file_iops = {
 	.setattr	= ocfs2_setattr,
 	.getattr	= ocfs2_getattr,
 	.permission	= ocfs2_permission,
+	.check_acl	= ocfs2_check_acl,
 };
 
 /*
diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index e5d738cd9cc0..33889dc52dd7 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -2498,4 +2498,5 @@ const struct inode_operations ocfs2_dir_iops = {
 	.listxattr	= ocfs2_listxattr,
 	.removexattr	= generic_removexattr,
 	.fiemap         = ocfs2_fiemap,
+	.check_acl	= ocfs2_check_acl,
 };
diff --git a/fs/proc/base.c b/fs/proc/base.c
index fc5bc2767692..8b8470113576 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -2169,7 +2169,7 @@ static const struct file_operations proc_fd_operations = {
  */
 static int proc_fd_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	int rv = generic_permission(inode, mask, flags, NULL);
+	int rv = generic_permission(inode, mask, flags);
 	if (rv == 0)
 		return 0;
 	if (task_pid(current) == proc_pid(inode))
diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c
index 91f080cc76c8..bbf31003d308 100644
--- a/fs/reiserfs/file.c
+++ b/fs/reiserfs/file.c
@@ -312,4 +312,5 @@ const struct inode_operations reiserfs_file_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
+	.check_acl = reiserfs_check_acl,
 };
diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c
index 118662690cdf..551f1b79dbc4 100644
--- a/fs/reiserfs/namei.c
+++ b/fs/reiserfs/namei.c
@@ -1529,6 +1529,7 @@ const struct inode_operations reiserfs_dir_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
+	.check_acl = reiserfs_check_acl,
 };
 
 /*
@@ -1545,6 +1546,7 @@ const struct inode_operations reiserfs_symlink_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
+	.check_acl = reiserfs_check_acl,
 
 };
 
@@ -1558,5 +1560,5 @@ const struct inode_operations reiserfs_special_inode_operations = {
 	.listxattr = reiserfs_listxattr,
 	.removexattr = reiserfs_removexattr,
 	.permission = reiserfs_permission,
-
+	.check_acl = reiserfs_check_acl,
 };
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index d78089690965..ddc5301d2986 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -868,11 +868,17 @@ out:
 	return err;
 }
 
-static int reiserfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int reiserfs_check_acl(struct inode *inode, int mask, unsigned int flags)
 {
 	struct posix_acl *acl;
 	int error = -EAGAIN; /* do regular unix permission checks by default */
 
+	/*
+	 * Stat data v1 doesn't support ACLs.
+	 */
+	if (get_inode_sd_version(inode) == STAT_DATA_V1)
+		return -EAGAIN;
+
 	if (flags & IPERM_FLAG_RCU)
 		return -ECHILD;
 
@@ -961,15 +967,7 @@ int reiserfs_permission(struct inode *inode, int mask, unsigned int flags)
 	if (IS_PRIVATE(inode))
 		return 0;
 
-#ifdef CONFIG_REISERFS_FS_XATTR
-	/*
-	 * Stat data v1 doesn't support ACLs.
-	 */
-	if (get_inode_sd_version(inode) != STAT_DATA_V1)
-		return generic_permission(inode, mask, flags,
-					reiserfs_check_acl);
-#endif
-	return generic_permission(inode, mask, flags, NULL);
+	return generic_permission(inode, mask, flags);
 }
 
 static int xattr_hide_revalidate(struct dentry *dentry, struct nameidata *nd)
diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c
index 0a12eb89cd32..a37165c64757 100644
--- a/fs/sysfs/inode.c
+++ b/fs/sysfs/inode.c
@@ -362,5 +362,5 @@ int sysfs_permission(struct inode *inode, int mask, unsigned int flags)
 	sysfs_refresh_inode(sd, inode);
 	mutex_unlock(&sysfs_mutex);
 
-	return generic_permission(inode, mask, flags, NULL);
+	return generic_permission(inode, mask, flags);
 }
-- 
cgit v1.2.3


From 1fc0f78ca9f311c6277e2f1b7655bb4d43ceb311 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 18:59:02 -0400
Subject: ->permission() sanitizing: MAY_NOT_BLOCK

Duplicate the flags argument into mask bitmap.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c            | 7 +++++--
 fs/proc/proc_sysctl.c | 2 +-
 2 files changed, 6 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index c5af0f37e67d..723a3fe4bc40 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -318,13 +318,16 @@ static inline int exec_permission(struct inode *inode, unsigned int flags)
 {
 	int ret;
 	struct user_namespace *ns = inode_userns(inode);
+	int mask = MAY_EXEC;
+	if (flags & IPERM_FLAG_RCU)
+		mask |= MAY_NOT_BLOCK;
 
 	if (inode->i_op->permission) {
-		ret = inode->i_op->permission(inode, MAY_EXEC, flags);
+		ret = inode->i_op->permission(inode, mask, flags);
 		if (likely(!ret))
 			goto ok;
 	} else {
-		ret = acl_permission_check(inode, MAY_EXEC, flags);
+		ret = acl_permission_check(inode, mask, flags);
 		if (likely(!ret))
 			goto ok;
 		if (ret != -EACCES)
diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index d167de365a8d..349b22f434d8 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -316,7 +316,7 @@ static int proc_sys_permission(struct inode *inode, int mask,unsigned int flags)
 	if (!table) /* global root - r-xr-xr-x */
 		error = mask & MAY_WRITE ? -EACCES : 0;
 	else /* Use the permissions on the sysctl table entry */
-		error = sysctl_perm(head->root, table, mask);
+		error = sysctl_perm(head->root, table, mask & ~MAY_NOT_BLOCK);
 
 	sysctl_head_finish(head);
 	return error;
-- 
cgit v1.2.3


From 9c2c703929e4c41210cfa6e3f599514421bab8dc Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 19:06:22 -0400
Subject: ->permission() sanitizing: pass MAY_NOT_BLOCK to ->check_acl()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/9p/acl.c                | 2 +-
 fs/btrfs/acl.c             | 3 +--
 fs/ext2/acl.c              | 2 +-
 fs/ext3/acl.c              | 2 +-
 fs/ext4/acl.c              | 2 +-
 fs/generic_acl.c           | 2 +-
 fs/gfs2/acl.c              | 2 +-
 fs/jffs2/acl.c             | 2 +-
 fs/jfs/acl.c               | 2 +-
 fs/namei.c                 | 4 ++--
 fs/ocfs2/acl.c             | 2 +-
 fs/reiserfs/xattr.c        | 2 +-
 fs/xfs/linux-2.6/xfs_acl.c | 2 +-
 13 files changed, 14 insertions(+), 15 deletions(-)

(limited to 'fs')

diff --git a/fs/9p/acl.c b/fs/9p/acl.c
index 535ab6eccb1a..94af68b092af 100644
--- a/fs/9p/acl.c
+++ b/fs/9p/acl.c
@@ -101,7 +101,7 @@ int v9fs_check_acl(struct inode *inode, int mask, unsigned int flags)
 	struct posix_acl *acl;
 	struct v9fs_session_info *v9ses;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	v9ses = v9fs_inode2v9ses(inode);
diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c
index f66fc9959733..a25a4a2e0df2 100644
--- a/fs/btrfs/acl.c
+++ b/fs/btrfs/acl.c
@@ -199,10 +199,9 @@ int btrfs_check_acl(struct inode *inode, int mask, unsigned int flags)
 {
 	int error = -EAGAIN;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			error = -ECHILD;
-
 	} else {
 		struct posix_acl *acl;
 		acl = btrfs_get_acl(inode, ACL_TYPE_ACCESS);
diff --git a/fs/ext2/acl.c b/fs/ext2/acl.c
index abea5a17c764..6b9442d1be52 100644
--- a/fs/ext2/acl.c
+++ b/fs/ext2/acl.c
@@ -236,7 +236,7 @@ ext2_check_acl(struct inode *inode, int mask, unsigned int flags)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c
index 9d021c0d472a..0a6940d6c30c 100644
--- a/fs/ext3/acl.c
+++ b/fs/ext3/acl.c
@@ -244,7 +244,7 @@ ext3_check_acl(struct inode *inode, int mask, unsigned int flags)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c
index 21eacd7b7d79..4f54252e439e 100644
--- a/fs/ext4/acl.c
+++ b/fs/ext4/acl.c
@@ -242,7 +242,7 @@ ext4_check_acl(struct inode *inode, int mask, unsigned int flags)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/generic_acl.c b/fs/generic_acl.c
index 8f26d1a58912..5976bb1fa4ca 100644
--- a/fs/generic_acl.c
+++ b/fs/generic_acl.c
@@ -192,7 +192,7 @@ generic_acl_chmod(struct inode *inode)
 int
 generic_check_acl(struct inode *inode, int mask, unsigned int flags)
 {
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 	} else {
diff --git a/fs/gfs2/acl.c b/fs/gfs2/acl.c
index cbc07155b1a0..4d97352d39a1 100644
--- a/fs/gfs2/acl.c
+++ b/fs/gfs2/acl.c
@@ -80,7 +80,7 @@ int gfs2_check_acl(struct inode *inode, int mask, unsigned int flags)
 	struct posix_acl *acl;
 	int error;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c
index 828a0e1ea438..952afb59e6f1 100644
--- a/fs/jffs2/acl.c
+++ b/fs/jffs2/acl.c
@@ -264,7 +264,7 @@ int jffs2_check_acl(struct inode *inode, int mask, unsigned int flags)
 	struct posix_acl *acl;
 	int rc;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	acl = jffs2_get_acl(inode, ACL_TYPE_ACCESS);
diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c
index e5de9422fa32..859ae5a92166 100644
--- a/fs/jfs/acl.c
+++ b/fs/jfs/acl.c
@@ -118,7 +118,7 @@ int jfs_check_acl(struct inode *inode, int mask, unsigned int flags)
 {
 	struct posix_acl *acl;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	acl = jfs_get_acl(inode, ACL_TYPE_ACCESS);
diff --git a/fs/namei.c b/fs/namei.c
index 723a3fe4bc40..e0624e2f0bba 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -181,7 +181,7 @@ static int acl_permission_check(struct inode *inode, int mask, unsigned int flag
 	int (*check_acl)(struct inode *inode, int mask, unsigned int flags);
 	unsigned int mode = inode->i_mode;
 
-	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
+	mask &= MAY_READ | MAY_WRITE | MAY_EXEC | MAY_NOT_BLOCK;
 
 	if (current_user_ns() != inode_userns(inode))
 		goto other_perms;
@@ -204,7 +204,7 @@ other_perms:
 	/*
 	 * If the DACs are ok we don't need any capability check.
 	 */
-	if ((mask & ~mode) == 0)
+	if ((mask & ~mode & (MAY_READ | MAY_WRITE | MAY_EXEC)) == 0)
 		return 0;
 	return -EACCES;
 }
diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c
index e913ad130fdd..4b683ccc4506 100644
--- a/fs/ocfs2/acl.c
+++ b/fs/ocfs2/acl.c
@@ -297,7 +297,7 @@ int ocfs2_check_acl(struct inode *inode, int mask, unsigned int flags)
 	struct posix_acl *acl;
 	int ret = -EAGAIN;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	osb = OCFS2_SB(inode->i_sb);
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index ddc5301d2986..6747470ec103 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -879,7 +879,7 @@ int reiserfs_check_acl(struct inode *inode, int mask, unsigned int flags)
 	if (get_inode_sd_version(inode) == STAT_DATA_V1)
 		return -EAGAIN;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	acl = reiserfs_get_acl(inode, ACL_TYPE_ACCESS);
diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c
index 39f4f809bb68..278e6736135a 100644
--- a/fs/xfs/linux-2.6/xfs_acl.c
+++ b/fs/xfs/linux-2.6/xfs_acl.c
@@ -235,7 +235,7 @@ xfs_check_acl(struct inode *inode, int mask, unsigned int flags)
 	if (!XFS_IFORK_Q(ip))
 		return -EAGAIN;
 
-	if (flags & IPERM_FLAG_RCU) {
+	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
 			return -ECHILD;
 		return -EAGAIN;
-- 
cgit v1.2.3


From 7e40145eb111a5192e6d819f764db9d6828d1abb Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 19:12:17 -0400
Subject: ->permission() sanitizing: don't pass flags to ->check_acl()

not used in the instances anymore.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/9p/acl.c                |  2 +-
 fs/9p/acl.h                |  2 +-
 fs/btrfs/acl.c             |  2 +-
 fs/btrfs/ctree.h           |  2 +-
 fs/ext2/acl.c              |  2 +-
 fs/ext2/acl.h              |  2 +-
 fs/ext3/acl.c              |  2 +-
 fs/ext3/acl.h              |  2 +-
 fs/ext4/acl.c              |  2 +-
 fs/ext4/acl.h              |  2 +-
 fs/generic_acl.c           |  2 +-
 fs/gfs2/acl.c              |  2 +-
 fs/gfs2/acl.h              |  2 +-
 fs/jffs2/acl.c             |  2 +-
 fs/jffs2/acl.h             |  2 +-
 fs/jfs/acl.c               |  2 +-
 fs/jfs/jfs_acl.h           |  2 +-
 fs/namei.c                 | 10 +++++-----
 fs/ocfs2/acl.c             |  2 +-
 fs/ocfs2/acl.h             |  2 +-
 fs/reiserfs/xattr.c        |  2 +-
 fs/xfs/linux-2.6/xfs_acl.c |  2 +-
 fs/xfs/xfs_acl.h           |  2 +-
 23 files changed, 27 insertions(+), 27 deletions(-)

(limited to 'fs')

diff --git a/fs/9p/acl.c b/fs/9p/acl.c
index 94af68b092af..e98f56d3787d 100644
--- a/fs/9p/acl.c
+++ b/fs/9p/acl.c
@@ -96,7 +96,7 @@ static struct posix_acl *v9fs_get_cached_acl(struct inode *inode, int type)
 	return acl;
 }
 
-int v9fs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int v9fs_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	struct v9fs_session_info *v9ses;
diff --git a/fs/9p/acl.h b/fs/9p/acl.h
index 7ef3ac9f6d95..59e18c2e8c7e 100644
--- a/fs/9p/acl.h
+++ b/fs/9p/acl.h
@@ -16,7 +16,7 @@
 
 #ifdef CONFIG_9P_FS_POSIX_ACL
 extern int v9fs_get_acl(struct inode *, struct p9_fid *);
-extern int v9fs_check_acl(struct inode *inode, int mask, unsigned int flags);
+extern int v9fs_check_acl(struct inode *inode, int mask);
 extern int v9fs_acl_chmod(struct dentry *);
 extern int v9fs_set_create_acl(struct dentry *,
 			       struct posix_acl *, struct posix_acl *);
diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c
index a25a4a2e0df2..9f62ab2a7282 100644
--- a/fs/btrfs/acl.c
+++ b/fs/btrfs/acl.c
@@ -195,7 +195,7 @@ out:
 	return ret;
 }
 
-int btrfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int btrfs_check_acl(struct inode *inode, int mask)
 {
 	int error = -EAGAIN;
 
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 3b859a3e6a0e..9864cec801ed 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2642,7 +2642,7 @@ do {								\
 
 /* acl.c */
 #ifdef CONFIG_BTRFS_FS_POSIX_ACL
-int btrfs_check_acl(struct inode *inode, int mask, unsigned int flags);
+int btrfs_check_acl(struct inode *inode, int mask);
 #else
 #define btrfs_check_acl NULL
 #endif
diff --git a/fs/ext2/acl.c b/fs/ext2/acl.c
index 6b9442d1be52..bfe651f9ae16 100644
--- a/fs/ext2/acl.c
+++ b/fs/ext2/acl.c
@@ -232,7 +232,7 @@ ext2_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 }
 
 int
-ext2_check_acl(struct inode *inode, int mask, unsigned int flags)
+ext2_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
diff --git a/fs/ext2/acl.h b/fs/ext2/acl.h
index c939b7b12099..3ff6cbb9ac44 100644
--- a/fs/ext2/acl.h
+++ b/fs/ext2/acl.h
@@ -54,7 +54,7 @@ static inline int ext2_acl_count(size_t size)
 #ifdef CONFIG_EXT2_FS_POSIX_ACL
 
 /* acl.c */
-extern int ext2_check_acl (struct inode *, int, unsigned int);
+extern int ext2_check_acl (struct inode *, int);
 extern int ext2_acl_chmod (struct inode *);
 extern int ext2_init_acl (struct inode *, struct inode *);
 
diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c
index 0a6940d6c30c..edfeb293d4cb 100644
--- a/fs/ext3/acl.c
+++ b/fs/ext3/acl.c
@@ -240,7 +240,7 @@ ext3_set_acl(handle_t *handle, struct inode *inode, int type,
 }
 
 int
-ext3_check_acl(struct inode *inode, int mask, unsigned int flags)
+ext3_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
diff --git a/fs/ext3/acl.h b/fs/ext3/acl.h
index 5faf8048e906..597334626de9 100644
--- a/fs/ext3/acl.h
+++ b/fs/ext3/acl.h
@@ -54,7 +54,7 @@ static inline int ext3_acl_count(size_t size)
 #ifdef CONFIG_EXT3_FS_POSIX_ACL
 
 /* acl.c */
-extern int ext3_check_acl (struct inode *, int, unsigned int);
+extern int ext3_check_acl (struct inode *, int);
 extern int ext3_acl_chmod (struct inode *);
 extern int ext3_init_acl (handle_t *, struct inode *, struct inode *);
 
diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c
index 4f54252e439e..60d900fcc3db 100644
--- a/fs/ext4/acl.c
+++ b/fs/ext4/acl.c
@@ -238,7 +238,7 @@ ext4_set_acl(handle_t *handle, struct inode *inode, int type,
 }
 
 int
-ext4_check_acl(struct inode *inode, int mask, unsigned int flags)
+ext4_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
diff --git a/fs/ext4/acl.h b/fs/ext4/acl.h
index dec821168fd4..9d843d5deac4 100644
--- a/fs/ext4/acl.h
+++ b/fs/ext4/acl.h
@@ -54,7 +54,7 @@ static inline int ext4_acl_count(size_t size)
 #ifdef CONFIG_EXT4_FS_POSIX_ACL
 
 /* acl.c */
-extern int ext4_check_acl(struct inode *, int, unsigned int);
+extern int ext4_check_acl(struct inode *, int);
 extern int ext4_acl_chmod(struct inode *);
 extern int ext4_init_acl(handle_t *, struct inode *, struct inode *);
 
diff --git a/fs/generic_acl.c b/fs/generic_acl.c
index 5976bb1fa4ca..70e90b4974ce 100644
--- a/fs/generic_acl.c
+++ b/fs/generic_acl.c
@@ -190,7 +190,7 @@ generic_acl_chmod(struct inode *inode)
 }
 
 int
-generic_check_acl(struct inode *inode, int mask, unsigned int flags)
+generic_check_acl(struct inode *inode, int mask)
 {
 	if (mask & MAY_NOT_BLOCK) {
 		if (!negative_cached_acl(inode, ACL_TYPE_ACCESS))
diff --git a/fs/gfs2/acl.c b/fs/gfs2/acl.c
index 4d97352d39a1..8ef1079f1665 100644
--- a/fs/gfs2/acl.c
+++ b/fs/gfs2/acl.c
@@ -75,7 +75,7 @@ static struct posix_acl *gfs2_acl_get(struct gfs2_inode *ip, int type)
  * Returns: errno
  */
 
-int gfs2_check_acl(struct inode *inode, int mask, unsigned int flags)
+int gfs2_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	int error;
diff --git a/fs/gfs2/acl.h b/fs/gfs2/acl.h
index a93907c8159b..b522b0cb39ea 100644
--- a/fs/gfs2/acl.h
+++ b/fs/gfs2/acl.h
@@ -16,7 +16,7 @@
 #define GFS2_POSIX_ACL_DEFAULT		"posix_acl_default"
 #define GFS2_ACL_MAX_ENTRIES		25
 
-extern int gfs2_check_acl(struct inode *inode, int mask, unsigned int);
+extern int gfs2_check_acl(struct inode *inode, int mask);
 extern int gfs2_acl_create(struct gfs2_inode *dip, struct inode *inode);
 extern int gfs2_acl_chmod(struct gfs2_inode *ip, struct iattr *attr);
 extern const struct xattr_handler gfs2_xattr_system_handler;
diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c
index 952afb59e6f1..3675b3cdee89 100644
--- a/fs/jffs2/acl.c
+++ b/fs/jffs2/acl.c
@@ -259,7 +259,7 @@ static int jffs2_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 	return rc;
 }
 
-int jffs2_check_acl(struct inode *inode, int mask, unsigned int flags)
+int jffs2_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	int rc;
diff --git a/fs/jffs2/acl.h b/fs/jffs2/acl.h
index 3119f59253d3..5e42de8d9541 100644
--- a/fs/jffs2/acl.h
+++ b/fs/jffs2/acl.h
@@ -26,7 +26,7 @@ struct jffs2_acl_header {
 
 #ifdef CONFIG_JFFS2_FS_POSIX_ACL
 
-extern int jffs2_check_acl(struct inode *, int, unsigned int);
+extern int jffs2_check_acl(struct inode *, int);
 extern int jffs2_acl_chmod(struct inode *);
 extern int jffs2_init_acl_pre(struct inode *, struct inode *, int *);
 extern int jffs2_init_acl_post(struct inode *);
diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c
index 859ae5a92166..8a0a0666d5a6 100644
--- a/fs/jfs/acl.c
+++ b/fs/jfs/acl.c
@@ -114,7 +114,7 @@ out:
 	return rc;
 }
 
-int jfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int jfs_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 
diff --git a/fs/jfs/jfs_acl.h b/fs/jfs/jfs_acl.h
index f9285c4900fa..54e07559878d 100644
--- a/fs/jfs/jfs_acl.h
+++ b/fs/jfs/jfs_acl.h
@@ -20,7 +20,7 @@
 
 #ifdef CONFIG_JFS_POSIX_ACL
 
-int jfs_check_acl(struct inode *, int, unsigned int flags);
+int jfs_check_acl(struct inode *, int);
 int jfs_init_acl(tid_t, struct inode *, struct inode *);
 int jfs_acl_chmod(struct inode *inode);
 
diff --git a/fs/namei.c b/fs/namei.c
index e0624e2f0bba..560fd1dff1d0 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -176,9 +176,9 @@ EXPORT_SYMBOL(putname);
 /*
  * This does basic POSIX ACL permission checking
  */
-static int acl_permission_check(struct inode *inode, int mask, unsigned int flags)
+static int acl_permission_check(struct inode *inode, int mask)
 {
-	int (*check_acl)(struct inode *inode, int mask, unsigned int flags);
+	int (*check_acl)(struct inode *inode, int mask);
 	unsigned int mode = inode->i_mode;
 
 	mask &= MAY_READ | MAY_WRITE | MAY_EXEC | MAY_NOT_BLOCK;
@@ -191,7 +191,7 @@ static int acl_permission_check(struct inode *inode, int mask, unsigned int flag
 	else {
 		check_acl = inode->i_op->check_acl;
 		if (IS_POSIXACL(inode) && (mode & S_IRWXG) && check_acl) {
-			int error = check_acl(inode, mask, flags);
+			int error = check_acl(inode, mask);
 			if (error != -EAGAIN)
 				return error;
 		}
@@ -231,7 +231,7 @@ int generic_permission(struct inode *inode, int mask, unsigned int flags)
 	/*
 	 * Do the basic POSIX ACL permission checks.
 	 */
-	ret = acl_permission_check(inode, mask, flags);
+	ret = acl_permission_check(inode, mask);
 	if (ret != -EACCES)
 		return ret;
 
@@ -327,7 +327,7 @@ static inline int exec_permission(struct inode *inode, unsigned int flags)
 		if (likely(!ret))
 			goto ok;
 	} else {
-		ret = acl_permission_check(inode, mask, flags);
+		ret = acl_permission_check(inode, mask);
 		if (likely(!ret))
 			goto ok;
 		if (ret != -EACCES)
diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c
index 4b683ccc4506..1cee970eb55a 100644
--- a/fs/ocfs2/acl.c
+++ b/fs/ocfs2/acl.c
@@ -290,7 +290,7 @@ static int ocfs2_set_acl(handle_t *handle,
 	return ret;
 }
 
-int ocfs2_check_acl(struct inode *inode, int mask, unsigned int flags)
+int ocfs2_check_acl(struct inode *inode, int mask)
 {
 	struct ocfs2_super *osb;
 	struct buffer_head *di_bh = NULL;
diff --git a/fs/ocfs2/acl.h b/fs/ocfs2/acl.h
index 4fe7c9cf4bfb..5c5d31f05853 100644
--- a/fs/ocfs2/acl.h
+++ b/fs/ocfs2/acl.h
@@ -26,7 +26,7 @@ struct ocfs2_acl_entry {
 	__le32 e_id;
 };
 
-extern int ocfs2_check_acl(struct inode *, int, unsigned int);
+extern int ocfs2_check_acl(struct inode *, int);
 extern int ocfs2_acl_chmod(struct inode *);
 extern int ocfs2_init_acl(handle_t *, struct inode *, struct inode *,
 			  struct buffer_head *, struct buffer_head *,
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index 6747470ec103..6ee3c11aa8d9 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -868,7 +868,7 @@ out:
 	return err;
 }
 
-int reiserfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+int reiserfs_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl;
 	int error = -EAGAIN; /* do regular unix permission checks by default */
diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c
index 278e6736135a..a5dcd6a0f1b5 100644
--- a/fs/xfs/linux-2.6/xfs_acl.c
+++ b/fs/xfs/linux-2.6/xfs_acl.c
@@ -219,7 +219,7 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl)
 }
 
 int
-xfs_check_acl(struct inode *inode, int mask, unsigned int flags)
+xfs_check_acl(struct inode *inode, int mask)
 {
 	struct xfs_inode *ip;
 	struct posix_acl *acl;
diff --git a/fs/xfs/xfs_acl.h b/fs/xfs/xfs_acl.h
index 11dd72070cbb..0135e2a669d7 100644
--- a/fs/xfs/xfs_acl.h
+++ b/fs/xfs/xfs_acl.h
@@ -42,7 +42,7 @@ struct xfs_acl {
 #define SGI_ACL_DEFAULT_SIZE	(sizeof(SGI_ACL_DEFAULT)-1)
 
 #ifdef CONFIG_XFS_POSIX_ACL
-extern int xfs_check_acl(struct inode *inode, int mask, unsigned int flags);
+extern int xfs_check_acl(struct inode *inode, int mask);
 extern struct posix_acl *xfs_get_acl(struct inode *inode, int type);
 extern int xfs_inherit_acl(struct inode *inode, struct posix_acl *default_acl);
 extern int xfs_acl_chmod(struct inode *inode);
-- 
cgit v1.2.3


From 2830ba7f34ebb27c4e5b8b6ef408cd6d74860890 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 19:16:29 -0400
Subject: ->permission() sanitizing: don't pass flags to generic_permission()

redundant; all callers get it duplicated in mask & MAY_NOT_BLOCK and none of
them removes that bit.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/afs/security.c       | 2 +-
 fs/btrfs/inode.c        | 2 +-
 fs/ceph/inode.c         | 2 +-
 fs/cifs/cifsfs.c        | 2 +-
 fs/fuse/dir.c           | 5 ++---
 fs/gfs2/inode.c         | 2 +-
 fs/hostfs/hostfs_kern.c | 2 +-
 fs/hpfs/namei.c         | 2 +-
 fs/namei.c              | 4 ++--
 fs/nfs/dir.c            | 2 +-
 fs/nilfs2/inode.c       | 2 +-
 fs/ocfs2/file.c         | 2 +-
 fs/proc/base.c          | 2 +-
 fs/reiserfs/xattr.c     | 2 +-
 fs/sysfs/inode.c        | 2 +-
 15 files changed, 17 insertions(+), 18 deletions(-)

(limited to 'fs')

diff --git a/fs/afs/security.c b/fs/afs/security.c
index 745ee654165f..ab6b3147f05f 100644
--- a/fs/afs/security.c
+++ b/fs/afs/security.c
@@ -350,7 +350,7 @@ int afs_permission(struct inode *inode, int mask, unsigned int flags)
 	}
 
 	key_put(key);
-	ret = generic_permission(inode, mask, flags);
+	ret = generic_permission(inode, mask);
 	_leave(" = %d", ret);
 	return ret;
 
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index f0bd87371566..b61b0477a8ac 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -7339,7 +7339,7 @@ static int btrfs_permission(struct inode *inode, int mask, unsigned int flags)
 		return -EROFS;
 	if ((BTRFS_I(inode)->flags & BTRFS_INODE_READONLY) && (mask & MAY_WRITE))
 		return -EACCES;
-	return generic_permission(inode, mask, flags);
+	return generic_permission(inode, mask);
 }
 
 static const struct inode_operations btrfs_dir_inode_operations = {
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index beb5d55d6fd2..9c169741b416 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -1805,7 +1805,7 @@ int ceph_permission(struct inode *inode, int mask, unsigned int flags)
 	err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED);
 
 	if (!err)
-		err = generic_permission(inode, mask, flags);
+		err = generic_permission(inode, mask);
 	return err;
 }
 
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index b79804fa410f..b285b5bcf711 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -239,7 +239,7 @@ static int cifs_permission(struct inode *inode, int mask, unsigned int flags)
 		on the client (above and beyond ACL on servers) for
 		servers which do not support setting and viewing mode bits,
 		so allowing client to check permissions is useful */
-		return generic_permission(inode, mask, flags);
+		return generic_permission(inode, mask);
 }
 
 static struct kmem_cache *cifs_inode_cachep;
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 0a2fcd860ad6..0df56b6c26e2 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -1018,7 +1018,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 	}
 
 	if (fc->flags & FUSE_DEFAULT_PERMISSIONS) {
-		err = generic_permission(inode, mask, flags);
+		err = generic_permission(inode, mask);
 
 		/* If permission is denied, try to refresh file
 		   attributes.  This is also needed, because the root
@@ -1026,8 +1026,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 		if (err == -EACCES && !refreshed) {
 			err = fuse_perm_getattr(inode, flags);
 			if (!err)
-				err = generic_permission(inode, mask,
-							flags);
+				err = generic_permission(inode, mask);
 		}
 
 		/* Note: the opposite of the above test does not
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index d5f0f4ea25dc..b776ec8f9c19 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -1564,7 +1564,7 @@ int gfs2_permission(struct inode *inode, int mask, unsigned int flags)
 	if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode))
 		error = -EACCES;
 	else
-		error = generic_permission(inode, mask, flags);
+		error = generic_permission(inode, mask);
 	if (unlock)
 		gfs2_glock_dq_uninit(&i_gh);
 
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index a98d0d1aef65..b1bc31bde833 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -770,7 +770,7 @@ int hostfs_permission(struct inode *ino, int desired, unsigned int flags)
 		err = access_file(name, r, w, x);
 	__putname(name);
 	if (!err)
-		err = generic_permission(ino, desired, flags);
+		err = generic_permission(ino, desired);
 	return err;
 }
 
diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c
index bd2ce7dd8df3..2df69e2f07cf 100644
--- a/fs/hpfs/namei.c
+++ b/fs/hpfs/namei.c
@@ -398,7 +398,7 @@ again:
 			hpfs_unlock(dir->i_sb);
 			return -ENOSPC;
 		}
-		if (generic_permission(inode, MAY_WRITE, 0) ||
+		if (generic_permission(inode, MAY_WRITE) ||
 		    !S_ISREG(inode->i_mode) ||
 		    get_write_access(inode)) {
 			d_rehash(dentry);
diff --git a/fs/namei.c b/fs/namei.c
index 560fd1dff1d0..684e0f30cf4c 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -224,7 +224,7 @@ other_perms:
  * request cannot be satisfied (eg. requires blocking or too much complexity).
  * It would then be called again in ref-walk mode.
  */
-int generic_permission(struct inode *inode, int mask, unsigned int flags)
+int generic_permission(struct inode *inode, int mask)
 {
 	int ret;
 
@@ -289,7 +289,7 @@ int inode_permission(struct inode *inode, int mask)
 	if (inode->i_op->permission)
 		retval = inode->i_op->permission(inode, mask, 0);
 	else
-		retval = generic_permission(inode, mask, 0);
+		retval = generic_permission(inode, mask);
 
 	if (retval)
 		return retval;
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 0485dca34fb1..16cf84b4afb9 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -2328,7 +2328,7 @@ out:
 out_notsup:
 	res = nfs_revalidate_inode(NFS_SERVER(inode), inode);
 	if (res == 0)
-		res = generic_permission(inode, mask, flags);
+		res = generic_permission(inode, mask);
 	goto out;
 }
 
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 650aa7755003..0df6de58bbcb 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -806,7 +806,7 @@ int nilfs_permission(struct inode *inode, int mask, unsigned int flags)
 	    root->cno != NILFS_CPTREE_CURRENT_CNO)
 		return -EROFS; /* snapshot is not writable */
 
-	return generic_permission(inode, mask, flags);
+	return generic_permission(inode, mask);
 }
 
 int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh)
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index d058cb7e12d4..ecb52b028964 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1293,7 +1293,7 @@ int ocfs2_permission(struct inode *inode, int mask, unsigned int flags)
 		goto out;
 	}
 
-	ret = generic_permission(inode, mask, flags);
+	ret = generic_permission(inode, mask);
 
 	ocfs2_inode_unlock(inode, 0);
 out:
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 8b8470113576..53a1a961b25f 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -2169,7 +2169,7 @@ static const struct file_operations proc_fd_operations = {
  */
 static int proc_fd_permission(struct inode *inode, int mask, unsigned int flags)
 {
-	int rv = generic_permission(inode, mask, flags);
+	int rv = generic_permission(inode, mask);
 	if (rv == 0)
 		return 0;
 	if (task_pid(current) == proc_pid(inode))
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index 6ee3c11aa8d9..f17319613a85 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -967,7 +967,7 @@ int reiserfs_permission(struct inode *inode, int mask, unsigned int flags)
 	if (IS_PRIVATE(inode))
 		return 0;
 
-	return generic_permission(inode, mask, flags);
+	return generic_permission(inode, mask);
 }
 
 static int xattr_hide_revalidate(struct dentry *dentry, struct nameidata *nd)
diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c
index a37165c64757..04c81e5ba6f7 100644
--- a/fs/sysfs/inode.c
+++ b/fs/sysfs/inode.c
@@ -362,5 +362,5 @@ int sysfs_permission(struct inode *inode, int mask, unsigned int flags)
 	sysfs_refresh_inode(sd, inode);
 	mutex_unlock(&sysfs_mutex);
 
-	return generic_permission(inode, mask, flags);
+	return generic_permission(inode, mask);
 }
-- 
cgit v1.2.3


From 10556cb21a0d0b24d95f00ea6df16f599a3345b2 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 19:28:19 -0400
Subject: ->permission() sanitizing: don't pass flags to ->permission()

not used by the instances anymore.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/afs/internal.h       |  2 +-
 fs/afs/security.c       |  4 ++--
 fs/bad_inode.c          |  2 +-
 fs/btrfs/inode.c        |  2 +-
 fs/ceph/inode.c         |  4 ++--
 fs/ceph/super.h         |  2 +-
 fs/cifs/cifsfs.c        |  2 +-
 fs/coda/coda_linux.h    |  2 +-
 fs/coda/dir.c           |  4 ++--
 fs/coda/pioctl.c        |  4 ++--
 fs/ecryptfs/inode.c     |  4 ++--
 fs/fuse/dir.c           | 14 +++++++-------
 fs/gfs2/file.c          |  2 +-
 fs/gfs2/inode.c         | 16 ++++++++--------
 fs/gfs2/inode.h         |  2 +-
 fs/hostfs/hostfs_kern.c |  4 ++--
 fs/namei.c              |  4 ++--
 fs/nfs/dir.c            |  4 ++--
 fs/nilfs2/inode.c       |  2 +-
 fs/nilfs2/nilfs.h       |  2 +-
 fs/ocfs2/file.c         |  4 ++--
 fs/ocfs2/file.h         |  2 +-
 fs/proc/base.c          |  2 +-
 fs/proc/proc_sysctl.c   |  2 +-
 fs/reiserfs/xattr.c     |  2 +-
 fs/sysfs/inode.c        |  4 ++--
 fs/sysfs/sysfs.h        |  2 +-
 27 files changed, 50 insertions(+), 50 deletions(-)

(limited to 'fs')

diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index 5a9b6843bac1..f396d337b817 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -627,7 +627,7 @@ extern void afs_clear_permits(struct afs_vnode *);
 extern void afs_cache_permit(struct afs_vnode *, struct key *, long);
 extern void afs_zap_permits(struct rcu_head *);
 extern struct key *afs_request_key(struct afs_cell *);
-extern int afs_permission(struct inode *, int, unsigned int);
+extern int afs_permission(struct inode *, int);
 
 /*
  * server.c
diff --git a/fs/afs/security.c b/fs/afs/security.c
index ab6b3147f05f..8d010422dc89 100644
--- a/fs/afs/security.c
+++ b/fs/afs/security.c
@@ -285,14 +285,14 @@ static int afs_check_permit(struct afs_vnode *vnode, struct key *key,
  * - AFS ACLs are attached to directories only, and a file is controlled by its
  *   parent directory's ACL
  */
-int afs_permission(struct inode *inode, int mask, unsigned int flags)
+int afs_permission(struct inode *inode, int mask)
 {
 	struct afs_vnode *vnode = AFS_FS_I(inode);
 	afs_access_t uninitialized_var(access);
 	struct key *key;
 	int ret;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	_enter("{{%x:%u},%lx},%x,",
diff --git a/fs/bad_inode.c b/fs/bad_inode.c
index bfcb18feb1df..f024d8aaddef 100644
--- a/fs/bad_inode.c
+++ b/fs/bad_inode.c
@@ -229,7 +229,7 @@ static int bad_inode_readlink(struct dentry *dentry, char __user *buffer,
 	return -EIO;
 }
 
-static int bad_inode_permission(struct inode *inode, int mask, unsigned int flags)
+static int bad_inode_permission(struct inode *inode, int mask)
 {
 	return -EIO;
 }
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index b61b0477a8ac..cb170ca51c64 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -7331,7 +7331,7 @@ static int btrfs_set_page_dirty(struct page *page)
 	return __set_page_dirty_nobuffers(page);
 }
 
-static int btrfs_permission(struct inode *inode, int mask, unsigned int flags)
+static int btrfs_permission(struct inode *inode, int mask)
 {
 	struct btrfs_root *root = BTRFS_I(inode)->root;
 
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 9c169741b416..dfb2831d8d85 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -1795,11 +1795,11 @@ int ceph_do_getattr(struct inode *inode, int mask)
  * Check inode permissions.  We verify we have a valid value for
  * the AUTH cap, then call the generic handler.
  */
-int ceph_permission(struct inode *inode, int mask, unsigned int flags)
+int ceph_permission(struct inode *inode, int mask)
 {
 	int err;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED);
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index f5cabefa98dc..56c41ef47cad 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -692,7 +692,7 @@ extern void ceph_queue_invalidate(struct inode *inode);
 extern void ceph_queue_writeback(struct inode *inode);
 
 extern int ceph_do_getattr(struct inode *inode, int mask);
-extern int ceph_permission(struct inode *inode, int mask, unsigned int flags);
+extern int ceph_permission(struct inode *inode, int mask);
 extern int ceph_setattr(struct dentry *dentry, struct iattr *attr);
 extern int ceph_getattr(struct vfsmount *mnt, struct dentry *dentry,
 			struct kstat *stat);
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index b285b5bcf711..cbbb55e781d3 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -224,7 +224,7 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	return 0;
 }
 
-static int cifs_permission(struct inode *inode, int mask, unsigned int flags)
+static int cifs_permission(struct inode *inode, int mask)
 {
 	struct cifs_sb_info *cifs_sb;
 
diff --git a/fs/coda/coda_linux.h b/fs/coda/coda_linux.h
index 9b0c5323890b..44e17e9c21ae 100644
--- a/fs/coda/coda_linux.h
+++ b/fs/coda/coda_linux.h
@@ -39,7 +39,7 @@ extern const struct file_operations coda_ioctl_operations;
 /* operations shared over more than one file */
 int coda_open(struct inode *i, struct file *f);
 int coda_release(struct inode *i, struct file *f);
-int coda_permission(struct inode *inode, int mask, unsigned int flags);
+int coda_permission(struct inode *inode, int mask);
 int coda_revalidate_inode(struct dentry *);
 int coda_getattr(struct vfsmount *, struct dentry *, struct kstat *);
 int coda_setattr(struct dentry *, struct iattr *);
diff --git a/fs/coda/dir.c b/fs/coda/dir.c
index 2b8dae4d121e..cd5532398c15 100644
--- a/fs/coda/dir.c
+++ b/fs/coda/dir.c
@@ -132,11 +132,11 @@ exit:
 }
 
 
-int coda_permission(struct inode *inode, int mask, unsigned int flags)
+int coda_permission(struct inode *inode, int mask)
 {
 	int error;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
diff --git a/fs/coda/pioctl.c b/fs/coda/pioctl.c
index cb140ef293e4..ee0981f1375b 100644
--- a/fs/coda/pioctl.c
+++ b/fs/coda/pioctl.c
@@ -24,7 +24,7 @@
 #include "coda_linux.h"
 
 /* pioctl ops */
-static int coda_ioctl_permission(struct inode *inode, int mask, unsigned int flags);
+static int coda_ioctl_permission(struct inode *inode, int mask);
 static long coda_pioctl(struct file *filp, unsigned int cmd,
 			unsigned long user_data);
 
@@ -41,7 +41,7 @@ const struct file_operations coda_ioctl_operations = {
 };
 
 /* the coda pioctl inode ops */
-static int coda_ioctl_permission(struct inode *inode, int mask, unsigned int flags)
+static int coda_ioctl_permission(struct inode *inode, int mask)
 {
 	return (mask & MAY_EXEC) ? -EACCES : 0;
 }
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 7349ade17de6..bec75f8e91ac 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -942,9 +942,9 @@ int ecryptfs_truncate(struct dentry *dentry, loff_t new_length)
 }
 
 static int
-ecryptfs_permission(struct inode *inode, int mask, unsigned int flags)
+ecryptfs_permission(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 	return inode_permission(ecryptfs_inode_to_lower(inode), mask);
 }
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 0df56b6c26e2..e2b14001cea5 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -971,9 +971,9 @@ static int fuse_access(struct inode *inode, int mask)
 	return err;
 }
 
-static int fuse_perm_getattr(struct inode *inode, int flags)
+static int fuse_perm_getattr(struct inode *inode, int mask)
 {
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	return fuse_do_getattr(inode, NULL, NULL);
@@ -992,7 +992,7 @@ static int fuse_perm_getattr(struct inode *inode, int flags)
  * access request is sent.  Execute permission is still checked
  * locally based on file mode.
  */
-static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
+static int fuse_permission(struct inode *inode, int mask)
 {
 	struct fuse_conn *fc = get_fuse_conn(inode);
 	bool refreshed = false;
@@ -1011,7 +1011,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 		if (fi->i_time < get_jiffies_64()) {
 			refreshed = true;
 
-			err = fuse_perm_getattr(inode, flags);
+			err = fuse_perm_getattr(inode, mask);
 			if (err)
 				return err;
 		}
@@ -1024,7 +1024,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 		   attributes.  This is also needed, because the root
 		   node will at first have no permissions */
 		if (err == -EACCES && !refreshed) {
-			err = fuse_perm_getattr(inode, flags);
+			err = fuse_perm_getattr(inode, mask);
 			if (!err)
 				err = generic_permission(inode, mask);
 		}
@@ -1034,7 +1034,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 		   noticed immediately, only after the attribute
 		   timeout has expired */
 	} else if (mask & (MAY_ACCESS | MAY_CHDIR)) {
-		if (flags & IPERM_FLAG_RCU)
+		if (mask & MAY_NOT_BLOCK)
 			return -ECHILD;
 
 		err = fuse_access(inode, mask);
@@ -1043,7 +1043,7 @@ static int fuse_permission(struct inode *inode, int mask, unsigned int flags)
 			if (refreshed)
 				return -EACCES;
 
-			err = fuse_perm_getattr(inode, flags);
+			err = fuse_perm_getattr(inode, mask);
 			if (!err && !(inode->i_mode & S_IXUGO))
 				return -EACCES;
 		}
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index a9f5cbe45cd9..89c39e53760d 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -243,7 +243,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask)
 	    !capable(CAP_LINUX_IMMUTABLE))
 		goto out;
 	if (!IS_IMMUTABLE(inode)) {
-		error = gfs2_permission(inode, MAY_WRITE, 0);
+		error = gfs2_permission(inode, MAY_WRITE);
 		if (error)
 			goto out;
 	}
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index b776ec8f9c19..b1090d66a6fd 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -307,7 +307,7 @@ struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name,
 	}
 
 	if (!is_root) {
-		error = gfs2_permission(dir, MAY_EXEC, 0);
+		error = gfs2_permission(dir, MAY_EXEC);
 		if (error)
 			goto out;
 	}
@@ -337,7 +337,7 @@ static int create_ok(struct gfs2_inode *dip, const struct qstr *name,
 {
 	int error;
 
-	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0);
+	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC);
 	if (error)
 		return error;
 
@@ -857,7 +857,7 @@ static int gfs2_link(struct dentry *old_dentry, struct inode *dir,
 	if (inode->i_nlink == 0)
 		goto out_gunlock;
 
-	error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC, 0);
+	error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC);
 	if (error)
 		goto out_gunlock;
 
@@ -990,7 +990,7 @@ static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name,
 	if (IS_APPEND(&dip->i_inode))
 		return -EPERM;
 
-	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0);
+	error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC);
 	if (error)
 		return error;
 
@@ -1336,7 +1336,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry,
 			}
 		}
 	} else {
-		error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC, 0);
+		error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC);
 		if (error)
 			goto out_gunlock;
 
@@ -1371,7 +1371,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry,
 	/* Check out the dir to be renamed */
 
 	if (dir_rename) {
-		error = gfs2_permission(odentry->d_inode, MAY_WRITE, 0);
+		error = gfs2_permission(odentry->d_inode, MAY_WRITE);
 		if (error)
 			goto out_gunlock;
 	}
@@ -1543,7 +1543,7 @@ static void gfs2_put_link(struct dentry *dentry, struct nameidata *nd, void *p)
  * Returns: errno
  */
 
-int gfs2_permission(struct inode *inode, int mask, unsigned int flags)
+int gfs2_permission(struct inode *inode, int mask)
 {
 	struct gfs2_inode *ip;
 	struct gfs2_holder i_gh;
@@ -1553,7 +1553,7 @@ int gfs2_permission(struct inode *inode, int mask, unsigned int flags)
 
 	ip = GFS2_I(inode);
 	if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) {
-		if (flags & IPERM_FLAG_RCU)
+		if (mask & MAY_NOT_BLOCK)
 			return -ECHILD;
 		error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh);
 		if (error)
diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h
index 31606076f701..8d90e0c07672 100644
--- a/fs/gfs2/inode.h
+++ b/fs/gfs2/inode.h
@@ -108,7 +108,7 @@ extern int gfs2_inode_refresh(struct gfs2_inode *ip);
 
 extern struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name,
 				  int is_root);
-extern int gfs2_permission(struct inode *inode, int mask, unsigned int flags);
+extern int gfs2_permission(struct inode *inode, int mask);
 extern int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr);
 extern struct inode *gfs2_lookup_simple(struct inode *dip, const char *name);
 extern void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf);
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index b1bc31bde833..6e449c599b9d 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -748,12 +748,12 @@ int hostfs_rename(struct inode *from_ino, struct dentry *from,
 	return err;
 }
 
-int hostfs_permission(struct inode *ino, int desired, unsigned int flags)
+int hostfs_permission(struct inode *ino, int desired)
 {
 	char *name;
 	int r = 0, w = 0, x = 0, err;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (desired & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	if (desired & MAY_READ) r = 1;
diff --git a/fs/namei.c b/fs/namei.c
index 684e0f30cf4c..c5c382620a86 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -287,7 +287,7 @@ int inode_permission(struct inode *inode, int mask)
 	}
 
 	if (inode->i_op->permission)
-		retval = inode->i_op->permission(inode, mask, 0);
+		retval = inode->i_op->permission(inode, mask);
 	else
 		retval = generic_permission(inode, mask);
 
@@ -323,7 +323,7 @@ static inline int exec_permission(struct inode *inode, unsigned int flags)
 		mask |= MAY_NOT_BLOCK;
 
 	if (inode->i_op->permission) {
-		ret = inode->i_op->permission(inode, mask, flags);
+		ret = inode->i_op->permission(inode, mask);
 		if (likely(!ret))
 			goto ok;
 	} else {
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 16cf84b4afb9..a86acd6d0cb4 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -2273,12 +2273,12 @@ int nfs_may_open(struct inode *inode, struct rpc_cred *cred, int openflags)
 	return nfs_do_access(inode, cred, nfs_open_permission_mask(openflags));
 }
 
-int nfs_permission(struct inode *inode, int mask, unsigned int flags)
+int nfs_permission(struct inode *inode, int mask)
 {
 	struct rpc_cred *cred;
 	int res = 0;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	nfs_inc_stats(inode, NFSIOS_VFSACCESS);
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 0df6de58bbcb..be8664c6a825 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -799,7 +799,7 @@ out_err:
 	return err;
 }
 
-int nilfs_permission(struct inode *inode, int mask, unsigned int flags)
+int nilfs_permission(struct inode *inode, int mask)
 {
 	struct nilfs_root *root = NILFS_I(inode)->i_root;
 	if ((mask & MAY_WRITE) && root &&
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index f02b9ad43a21..6fb7511c0328 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -264,7 +264,7 @@ extern void nilfs_update_inode(struct inode *, struct buffer_head *);
 extern void nilfs_truncate(struct inode *);
 extern void nilfs_evict_inode(struct inode *);
 extern int nilfs_setattr(struct dentry *, struct iattr *);
-int nilfs_permission(struct inode *inode, int mask, unsigned int flags);
+int nilfs_permission(struct inode *inode, int mask);
 int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh);
 extern int nilfs_inode_dirty(struct inode *);
 int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty);
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index ecb52b028964..1406c37a5722 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1279,11 +1279,11 @@ bail:
 	return err;
 }
 
-int ocfs2_permission(struct inode *inode, int mask, unsigned int flags)
+int ocfs2_permission(struct inode *inode, int mask)
 {
 	int ret;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	ret = ocfs2_inode_lock(inode, NULL, 0);
diff --git a/fs/ocfs2/file.h b/fs/ocfs2/file.h
index f5afbbef6703..97bf761c9e7c 100644
--- a/fs/ocfs2/file.h
+++ b/fs/ocfs2/file.h
@@ -61,7 +61,7 @@ int ocfs2_zero_extend(struct inode *inode, struct buffer_head *di_bh,
 int ocfs2_setattr(struct dentry *dentry, struct iattr *attr);
 int ocfs2_getattr(struct vfsmount *mnt, struct dentry *dentry,
 		  struct kstat *stat);
-int ocfs2_permission(struct inode *inode, int mask, unsigned int flags);
+int ocfs2_permission(struct inode *inode, int mask);
 
 int ocfs2_should_update_atime(struct inode *inode,
 			      struct vfsmount *vfsmnt);
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 53a1a961b25f..be1ff932033b 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -2167,7 +2167,7 @@ static const struct file_operations proc_fd_operations = {
  * /proc/pid/fd needs a special permission handler so that a process can still
  * access /proc/self/fd after it has executed a setuid().
  */
-static int proc_fd_permission(struct inode *inode, int mask, unsigned int flags)
+static int proc_fd_permission(struct inode *inode, int mask)
 {
 	int rv = generic_permission(inode, mask);
 	if (rv == 0)
diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index 349b22f434d8..1a77dbef226f 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -294,7 +294,7 @@ out:
 	return ret;
 }
 
-static int proc_sys_permission(struct inode *inode, int mask,unsigned int flags)
+static int proc_sys_permission(struct inode *inode, int mask)
 {
 	/*
 	 * sysctl entries that are not writeable,
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index f17319613a85..4ea2ab41fdee 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -958,7 +958,7 @@ static int xattr_mount_check(struct super_block *s)
 	return 0;
 }
 
-int reiserfs_permission(struct inode *inode, int mask, unsigned int flags)
+int reiserfs_permission(struct inode *inode, int mask)
 {
 	/*
 	 * We don't do permission checks on the internal objects.
diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c
index 04c81e5ba6f7..e3f091a81c72 100644
--- a/fs/sysfs/inode.c
+++ b/fs/sysfs/inode.c
@@ -349,11 +349,11 @@ int sysfs_hash_and_remove(struct sysfs_dirent *dir_sd, const void *ns, const cha
 		return -ENOENT;
 }
 
-int sysfs_permission(struct inode *inode, int mask, unsigned int flags)
+int sysfs_permission(struct inode *inode, int mask)
 {
 	struct sysfs_dirent *sd;
 
-	if (flags & IPERM_FLAG_RCU)
+	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	sd = inode->i_private;
diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h
index 2ed2404f3113..845ab3ad229d 100644
--- a/fs/sysfs/sysfs.h
+++ b/fs/sysfs/sysfs.h
@@ -201,7 +201,7 @@ static inline void __sysfs_put(struct sysfs_dirent *sd)
 struct inode *sysfs_get_inode(struct super_block *sb, struct sysfs_dirent *sd);
 void sysfs_evict_inode(struct inode *inode);
 int sysfs_sd_setattr(struct sysfs_dirent *sd, struct iattr *iattr);
-int sysfs_permission(struct inode *inode, int mask, unsigned int flags);
+int sysfs_permission(struct inode *inode, int mask);
 int sysfs_setattr(struct dentry *dentry, struct iattr *iattr);
 int sysfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat);
 int sysfs_setxattr(struct dentry *dentry, const char *name, const void *value,
-- 
cgit v1.2.3


From eecdd358b467405a084d400d5ec571bbdbfe97a3 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 19:48:41 -0400
Subject: ->permission() sanitizing: don't pass flags to exec_permission()

pass mask instead; kill security_inode_exec_permission() since we can use
security_inode_permission() instead.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index c5c382620a86..21eba95368f2 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -304,7 +304,7 @@ int inode_permission(struct inode *inode, int mask)
 /**
  * exec_permission  -  check for right to do lookups in a given directory
  * @inode:	inode to check permission on
- * @flags:	IPERM_FLAG_ flags.
+ * @mask:	MAY_EXEC and possibly MAY_NOT_BLOCK flags.
  *
  * Short-cut version of inode_permission(), for calling on directories
  * during pathname resolution.  Combines parts of inode_permission()
@@ -314,13 +314,10 @@ int inode_permission(struct inode *inode, int mask)
  * short-cut DAC fails, then call ->permission() to do more
  * complete permission check.
  */
-static inline int exec_permission(struct inode *inode, unsigned int flags)
+static inline int exec_permission(struct inode *inode, int mask)
 {
 	int ret;
 	struct user_namespace *ns = inode_userns(inode);
-	int mask = MAY_EXEC;
-	if (flags & IPERM_FLAG_RCU)
-		mask |= MAY_NOT_BLOCK;
 
 	if (inode->i_op->permission) {
 		ret = inode->i_op->permission(inode, mask);
@@ -338,7 +335,7 @@ static inline int exec_permission(struct inode *inode, unsigned int flags)
 	}
 	return ret;
 ok:
-	return security_inode_exec_permission(inode, flags);
+	return security_inode_permission(inode, mask);
 }
 
 /**
@@ -1214,13 +1211,13 @@ retry:
 static inline int may_lookup(struct nameidata *nd)
 {
 	if (nd->flags & LOOKUP_RCU) {
-		int err = exec_permission(nd->inode, IPERM_FLAG_RCU);
+		int err = exec_permission(nd->inode, MAY_EXEC|MAY_NOT_BLOCK);
 		if (err != -ECHILD)
 			return err;
 		if (unlazy_walk(nd, NULL))
 			return -ECHILD;
 	}
-	return exec_permission(nd->inode, 0);
+	return exec_permission(nd->inode, MAY_EXEC);
 }
 
 static inline int handle_dots(struct nameidata *nd, int type)
@@ -1495,7 +1492,7 @@ static int path_init(int dfd, const char *name, unsigned int flags,
 			if (!S_ISDIR(dentry->d_inode->i_mode))
 				goto fput_fail;
 
-			retval = exec_permission(dentry->d_inode, 0);
+			retval = exec_permission(dentry->d_inode, MAY_EXEC);
 			if (retval)
 				goto fput_fail;
 		}
@@ -1652,7 +1649,7 @@ static struct dentry *__lookup_hash(struct qstr *name,
 	struct dentry *dentry;
 	int err;
 
-	err = exec_permission(inode, 0);
+	err = exec_permission(inode, MAY_EXEC);
 	if (err)
 		return ERR_PTR(err);
 
-- 
cgit v1.2.3


From d594e7ec4da6c1a527dea4965eee37cd785792b8 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 19:55:42 -0400
Subject: massage generic_permission() to treat directories on a separate path

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 21eba95368f2..758bae739305 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -235,12 +235,21 @@ int generic_permission(struct inode *inode, int mask)
 	if (ret != -EACCES)
 		return ret;
 
+	if (S_ISDIR(inode->i_mode)) {
+		/* DACs are overridable for directories */
+		if (ns_capable(inode_userns(inode), CAP_DAC_OVERRIDE))
+			return 0;
+		if (!(mask & MAY_WRITE))
+			if (ns_capable(inode_userns(inode), CAP_DAC_READ_SEARCH))
+				return 0;
+		return -EACCES;
+	}
 	/*
 	 * Read/write DACs are always overridable.
-	 * Executable DACs are overridable for all directories and
-	 * for non-directories that have least one exec bit set.
+	 * Executable DACs are overridable when there is
+	 * at least one exec bit set.
 	 */
-	if (!(mask & MAY_EXEC) || execute_ok(inode))
+	if (!(mask & MAY_EXEC) || (inode->i_mode & S_IXUGO))
 		if (ns_capable(inode_userns(inode), CAP_DAC_OVERRIDE))
 			return 0;
 
@@ -248,7 +257,7 @@ int generic_permission(struct inode *inode, int mask)
 	 * Searching includes executable on directories, else just read.
 	 */
 	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
-	if (mask == MAY_READ || (S_ISDIR(inode->i_mode) && !(mask & MAY_WRITE)))
+	if (mask == MAY_READ)
 		if (ns_capable(inode_userns(inode), CAP_DAC_READ_SEARCH))
 			return 0;
 
-- 
cgit v1.2.3


From 4ad5abb3d01a2c10854969b00982fadb130784a6 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 19:57:03 -0400
Subject: no reason to keep exec_permission() separate now

cache footprint alone makes it a bad idea...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 45 ++++-----------------------------------------
 1 file changed, 4 insertions(+), 41 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 758bae739305..b3c1a975c834 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -310,43 +310,6 @@ int inode_permission(struct inode *inode, int mask)
 	return security_inode_permission(inode, mask);
 }
 
-/**
- * exec_permission  -  check for right to do lookups in a given directory
- * @inode:	inode to check permission on
- * @mask:	MAY_EXEC and possibly MAY_NOT_BLOCK flags.
- *
- * Short-cut version of inode_permission(), for calling on directories
- * during pathname resolution.  Combines parts of inode_permission()
- * and generic_permission(), and tests ONLY for MAY_EXEC permission.
- *
- * If appropriate, check DAC only.  If not appropriate, or
- * short-cut DAC fails, then call ->permission() to do more
- * complete permission check.
- */
-static inline int exec_permission(struct inode *inode, int mask)
-{
-	int ret;
-	struct user_namespace *ns = inode_userns(inode);
-
-	if (inode->i_op->permission) {
-		ret = inode->i_op->permission(inode, mask);
-		if (likely(!ret))
-			goto ok;
-	} else {
-		ret = acl_permission_check(inode, mask);
-		if (likely(!ret))
-			goto ok;
-		if (ret != -EACCES)
-			return ret;
-		if (ns_capable(ns, CAP_DAC_OVERRIDE) ||
-				ns_capable(ns, CAP_DAC_READ_SEARCH))
-			goto ok;
-	}
-	return ret;
-ok:
-	return security_inode_permission(inode, mask);
-}
-
 /**
  * path_get - get a reference to a path
  * @path: path to get the reference to
@@ -1220,13 +1183,13 @@ retry:
 static inline int may_lookup(struct nameidata *nd)
 {
 	if (nd->flags & LOOKUP_RCU) {
-		int err = exec_permission(nd->inode, MAY_EXEC|MAY_NOT_BLOCK);
+		int err = inode_permission(nd->inode, MAY_EXEC|MAY_NOT_BLOCK);
 		if (err != -ECHILD)
 			return err;
 		if (unlazy_walk(nd, NULL))
 			return -ECHILD;
 	}
-	return exec_permission(nd->inode, MAY_EXEC);
+	return inode_permission(nd->inode, MAY_EXEC);
 }
 
 static inline int handle_dots(struct nameidata *nd, int type)
@@ -1501,7 +1464,7 @@ static int path_init(int dfd, const char *name, unsigned int flags,
 			if (!S_ISDIR(dentry->d_inode->i_mode))
 				goto fput_fail;
 
-			retval = exec_permission(dentry->d_inode, MAY_EXEC);
+			retval = inode_permission(dentry->d_inode, MAY_EXEC);
 			if (retval)
 				goto fput_fail;
 		}
@@ -1658,7 +1621,7 @@ static struct dentry *__lookup_hash(struct qstr *name,
 	struct dentry *dentry;
 	int err;
 
-	err = exec_permission(inode, MAY_EXEC);
+	err = inode_permission(inode, MAY_EXEC);
 	if (err)
 		return ERR_PTR(err);
 
-- 
cgit v1.2.3


From d2d9e9fbc2f8f492dae373482da61d34475c53c1 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 20 Jun 2011 10:55:26 -0400
Subject: merge do_revalidate() into its only caller

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 42 ++++++++++++++++++------------------------
 1 file changed, 18 insertions(+), 24 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index b3c1a975c834..42ccb97cc260 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -436,28 +436,6 @@ static inline int d_revalidate(struct dentry *dentry, struct nameidata *nd)
 	return dentry->d_op->d_revalidate(dentry, nd);
 }
 
-static struct dentry *
-do_revalidate(struct dentry *dentry, struct nameidata *nd)
-{
-	int status = d_revalidate(dentry, nd);
-	if (unlikely(status <= 0)) {
-		/*
-		 * The dentry failed validation.
-		 * If d_revalidate returned 0 attempt to invalidate
-		 * the dentry otherwise d_revalidate is asking us
-		 * to return a fail status.
-		 */
-		if (status < 0) {
-			dput(dentry);
-			dentry = ERR_PTR(status);
-		} else if (!d_invalidate(dentry)) {
-			dput(dentry);
-			dentry = NULL;
-		}
-	}
-	return dentry;
-}
-
 /**
  * complete_walk - successful completion of path walk
  * @nd:  pointer nameidata
@@ -1642,8 +1620,24 @@ static struct dentry *__lookup_hash(struct qstr *name,
 			return dentry;
 	}
 
-	if (dentry && (dentry->d_flags & DCACHE_OP_REVALIDATE))
-		dentry = do_revalidate(dentry, nd);
+	if (dentry && (dentry->d_flags & DCACHE_OP_REVALIDATE)) {
+		int status = d_revalidate(dentry, nd);
+		if (unlikely(status <= 0)) {
+			/*
+			 * The dentry failed validation.
+			 * If d_revalidate returned 0 attempt to invalidate
+			 * the dentry otherwise d_revalidate is asking us
+			 * to return a fail status.
+			 */
+			if (status < 0) {
+				dput(dentry);
+				return ERR_PTR(status);
+			} else if (!d_invalidate(dentry)) {
+				dput(dentry);
+				dentry = NULL;
+			}
+		}
+	}
 
 	if (!dentry)
 		dentry = d_alloc_and_lookup(base, name, nd);
-- 
cgit v1.2.3


From beefebf1aa611242e022e71bae87034f415d3314 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Tue, 21 Jun 2011 01:01:59 -0400
Subject: ecryptfs_inode_permission() doesn't need to bail out on RCU

... now that inode_permission() can take MAY_NOT_BLOCK and handle it
properly.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ecryptfs/inode.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'fs')

diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index bec75f8e91ac..799e01055f22 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -944,8 +944,6 @@ int ecryptfs_truncate(struct dentry *dentry, loff_t new_length)
 static int
 ecryptfs_permission(struct inode *inode, int mask)
 {
-	if (mask & MAY_NOT_BLOCK)
-		return -ECHILD;
 	return inode_permission(ecryptfs_inode_to_lower(inode), mask);
 }
 
-- 
cgit v1.2.3


From 7c97c200e2c5aa8b1067bebb99df0a7c2e12ebf3 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Tue, 21 Jun 2011 08:51:28 -0400
Subject: cifs: fix the type of cifs_demultiplex_thread()

... and get rid of a bogus typecast, while we are at it; it's not
just that we want a function returning int and not void, but cast
to pointer to function taking void * and returning void would be
(void (*)(void *)) and not (void *)(void *), TYVM...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/cifs/connect.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index ccc1afa0bf3b..e66297bad412 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -320,9 +320,10 @@ requeue_echo:
 }
 
 static int
-cifs_demultiplex_thread(struct TCP_Server_Info *server)
+cifs_demultiplex_thread(void *p)
 {
 	int length;
+	struct TCP_Server_Info *server = p;
 	unsigned int pdu_length, total_read;
 	struct smb_hdr *smb_buffer = NULL;
 	struct smb_hdr *bigbuf = NULL;
@@ -1791,7 +1792,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info)
 	 * this will succeed. No need for try_module_get().
 	 */
 	__module_get(THIS_MODULE);
-	tcp_ses->tsk = kthread_run((void *)(void *)cifs_demultiplex_thread,
+	tcp_ses->tsk = kthread_run(cifs_demultiplex_thread,
 				  tcp_ses, "cifsd");
 	if (IS_ERR(tcp_ses->tsk)) {
 		rc = PTR_ERR(tcp_ses->tsk);
-- 
cgit v1.2.3


From 643168c2dc0d04cb884ae7bcd1ee19d9ecdff6a8 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Wed, 22 Jun 2011 18:20:23 -0400
Subject: nfs4_closedata doesn't need to mess with struct path

instead of path_get()/path_put(), we can just use nfs_sb_{,de}active()
to pin the superblock down.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfs/nfs4_fs.h   | 10 +++++-----
 fs/nfs/nfs4proc.c  | 21 ++++++++++-----------
 fs/nfs/nfs4state.c | 12 ++++++------
 3 files changed, 21 insertions(+), 22 deletions(-)

(limited to 'fs')

diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index c4a69833dd0d..b788f2eb1ba0 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -238,7 +238,7 @@ extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *);
 extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *);
 extern int nfs4_init_clientid(struct nfs_client *, struct rpc_cred *);
 extern int nfs41_init_clientid(struct nfs_client *, struct rpc_cred *);
-extern int nfs4_do_close(struct path *path, struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc);
+extern int nfs4_do_close(struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc);
 extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle);
 extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name,
 		struct nfs4_fs_locations *fs_locations, struct page *page);
@@ -341,8 +341,8 @@ extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struc
 extern void nfs4_put_state_owner(struct nfs4_state_owner *);
 extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state_owner *);
 extern void nfs4_put_open_state(struct nfs4_state *);
-extern void nfs4_close_state(struct path *, struct nfs4_state *, fmode_t);
-extern void nfs4_close_sync(struct path *, struct nfs4_state *, fmode_t);
+extern void nfs4_close_state(struct nfs4_state *, fmode_t);
+extern void nfs4_close_sync(struct nfs4_state *, fmode_t);
 extern void nfs4_state_set_mode_locked(struct nfs4_state *, fmode_t);
 extern void nfs4_schedule_lease_recovery(struct nfs_client *);
 extern void nfs4_schedule_state_manager(struct nfs_client *);
@@ -373,8 +373,8 @@ extern struct svc_version nfs4_callback_version4;
 
 #else
 
-#define nfs4_close_state(a, b, c) do { } while (0)
-#define nfs4_close_sync(a, b, c) do { } while (0)
+#define nfs4_close_state(a, b) do { } while (0)
+#define nfs4_close_sync(a, b) do { } while (0)
 
 #endif /* CONFIG_NFS_V4 */
 #endif /* __LINUX_FS_NFS_NFS4_FS.H */
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 5879b23e0c99..e99d674149ae 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1154,7 +1154,7 @@ static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, fmode_t fmod
 	newstate = nfs4_opendata_to_nfs4_state(opendata);
 	if (IS_ERR(newstate))
 		return PTR_ERR(newstate);
-	nfs4_close_state(&opendata->path, newstate, fmode);
+	nfs4_close_state(newstate, fmode);
 	*res = newstate;
 	return 0;
 }
@@ -1352,7 +1352,7 @@ static void nfs4_open_confirm_release(void *calldata)
 		goto out_free;
 	state = nfs4_opendata_to_nfs4_state(data);
 	if (!IS_ERR(state))
-		nfs4_close_state(&data->path, state, data->o_arg.fmode);
+		nfs4_close_state(state, data->o_arg.fmode);
 out_free:
 	nfs4_opendata_put(data);
 }
@@ -1497,7 +1497,7 @@ static void nfs4_open_release(void *calldata)
 		goto out_free;
 	state = nfs4_opendata_to_nfs4_state(data);
 	if (!IS_ERR(state))
-		nfs4_close_state(&data->path, state, data->o_arg.fmode);
+		nfs4_close_state(state, data->o_arg.fmode);
 out_free:
 	nfs4_opendata_put(data);
 }
@@ -1873,7 +1873,6 @@ static int nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 }
 
 struct nfs4_closedata {
-	struct path path;
 	struct inode *inode;
 	struct nfs4_state *state;
 	struct nfs_closeargs arg;
@@ -1888,13 +1887,14 @@ static void nfs4_free_closedata(void *data)
 {
 	struct nfs4_closedata *calldata = data;
 	struct nfs4_state_owner *sp = calldata->state->owner;
+	struct super_block *sb = calldata->state->inode->i_sb;
 
 	if (calldata->roc)
 		pnfs_roc_release(calldata->state->inode);
 	nfs4_put_open_state(calldata->state);
 	nfs_free_seqid(calldata->arg.seqid);
 	nfs4_put_state_owner(sp);
-	path_put(&calldata->path);
+	nfs_sb_deactive(sb);
 	kfree(calldata);
 }
 
@@ -2014,7 +2014,7 @@ static const struct rpc_call_ops nfs4_close_ops = {
  *
  * NOTE: Caller must be holding the sp->so_owner semaphore!
  */
-int nfs4_do_close(struct path *path, struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc)
+int nfs4_do_close(struct nfs4_state *state, gfp_t gfp_mask, int wait, bool roc)
 {
 	struct nfs_server *server = NFS_SERVER(state->inode);
 	struct nfs4_closedata *calldata;
@@ -2050,8 +2050,7 @@ int nfs4_do_close(struct path *path, struct nfs4_state *state, gfp_t gfp_mask, i
 	calldata->res.seqid = calldata->arg.seqid;
 	calldata->res.server = server;
 	calldata->roc = roc;
-	path_get(path);
-	calldata->path = *path;
+	nfs_sb_active(calldata->inode->i_sb);
 
 	msg.rpc_argp = &calldata->arg;
 	msg.rpc_resp = &calldata->res;
@@ -2092,9 +2091,9 @@ static void nfs4_close_context(struct nfs_open_context *ctx, int is_sync)
 	if (ctx->state == NULL)
 		return;
 	if (is_sync)
-		nfs4_close_sync(&ctx->path, ctx->state, ctx->mode);
+		nfs4_close_sync(ctx->state, ctx->mode);
 	else
-		nfs4_close_state(&ctx->path, ctx->state, ctx->mode);
+		nfs4_close_state(ctx->state, ctx->mode);
 }
 
 static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle)
@@ -2642,7 +2641,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
 	if (ctx != NULL)
 		ctx->state = state;
 	else
-		nfs4_close_sync(path, state, fmode);
+		nfs4_close_sync(state, fmode);
 out:
 	return status;
 }
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index e97dd219f84f..7acfe8843626 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -641,7 +641,7 @@ void nfs4_put_open_state(struct nfs4_state *state)
 /*
  * Close the current file.
  */
-static void __nfs4_close(struct path *path, struct nfs4_state *state,
+static void __nfs4_close(struct nfs4_state *state,
 		fmode_t fmode, gfp_t gfp_mask, int wait)
 {
 	struct nfs4_state_owner *owner = state->owner;
@@ -685,18 +685,18 @@ static void __nfs4_close(struct path *path, struct nfs4_state *state,
 	} else {
 		bool roc = pnfs_roc(state->inode);
 
-		nfs4_do_close(path, state, gfp_mask, wait, roc);
+		nfs4_do_close(state, gfp_mask, wait, roc);
 	}
 }
 
-void nfs4_close_state(struct path *path, struct nfs4_state *state, fmode_t fmode)
+void nfs4_close_state(struct nfs4_state *state, fmode_t fmode)
 {
-	__nfs4_close(path, state, fmode, GFP_NOFS, 0);
+	__nfs4_close(state, fmode, GFP_NOFS, 0);
 }
 
-void nfs4_close_sync(struct path *path, struct nfs4_state *state, fmode_t fmode)
+void nfs4_close_sync(struct nfs4_state *state, fmode_t fmode)
 {
-	__nfs4_close(path, state, fmode, GFP_KERNEL, 1);
+	__nfs4_close(state, fmode, GFP_KERNEL, 1);
 }
 
 /*
-- 
cgit v1.2.3


From 82a2c1b77ada3b81911deeb871719ec5ff77f83c Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Wed, 22 Jun 2011 18:30:55 -0400
Subject: nfs4_opendata doesn't need struct path either

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfs/nfs4proc.c | 45 ++++++++++++++++++++++-----------------------
 1 file changed, 22 insertions(+), 23 deletions(-)

(limited to 'fs')

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index e99d674149ae..aef22fa0a1c0 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -763,8 +763,8 @@ struct nfs4_opendata {
 	struct nfs_open_confirmres c_res;
 	struct nfs_fattr f_attr;
 	struct nfs_fattr dir_attr;
-	struct path path;
 	struct dentry *dir;
+	struct dentry *dentry;
 	struct nfs4_state_owner *owner;
 	struct nfs4_state *state;
 	struct iattr attrs;
@@ -786,12 +786,12 @@ static void nfs4_init_opendata_res(struct nfs4_opendata *p)
 	nfs_fattr_init(&p->dir_attr);
 }
 
-static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
+static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry,
 		struct nfs4_state_owner *sp, fmode_t fmode, int flags,
 		const struct iattr *attrs,
 		gfp_t gfp_mask)
 {
-	struct dentry *parent = dget_parent(path->dentry);
+	struct dentry *parent = dget_parent(dentry);
 	struct inode *dir = parent->d_inode;
 	struct nfs_server *server = NFS_SERVER(dir);
 	struct nfs4_opendata *p;
@@ -802,8 +802,8 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
 	p->o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid, gfp_mask);
 	if (p->o_arg.seqid == NULL)
 		goto err_free;
-	path_get(path);
-	p->path = *path;
+	nfs_sb_active(dentry->d_sb);
+	p->dentry = dget(dentry);
 	p->dir = parent;
 	p->owner = sp;
 	atomic_inc(&sp->so_count);
@@ -812,7 +812,7 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path,
 	p->o_arg.fmode = fmode & (FMODE_READ|FMODE_WRITE);
 	p->o_arg.clientid = server->nfs_client->cl_clientid;
 	p->o_arg.id = sp->so_owner_id.id;
-	p->o_arg.name = &p->path.dentry->d_name;
+	p->o_arg.name = &dentry->d_name;
 	p->o_arg.server = server;
 	p->o_arg.bitmask = server->attr_bitmask;
 	p->o_arg.claim = NFS4_OPEN_CLAIM_NULL;
@@ -842,13 +842,15 @@ static void nfs4_opendata_free(struct kref *kref)
 {
 	struct nfs4_opendata *p = container_of(kref,
 			struct nfs4_opendata, kref);
+	struct super_block *sb = p->dentry->d_sb;
 
 	nfs_free_seqid(p->o_arg.seqid);
 	if (p->state != NULL)
 		nfs4_put_open_state(p->state);
 	nfs4_put_state_owner(p->owner);
 	dput(p->dir);
-	path_put(&p->path);
+	dput(p->dentry);
+	nfs_sb_deactive(sb);
 	kfree(p);
 }
 
@@ -1130,7 +1132,7 @@ static struct nfs4_opendata *nfs4_open_recoverdata_alloc(struct nfs_open_context
 {
 	struct nfs4_opendata *opendata;
 
-	opendata = nfs4_opendata_alloc(&ctx->path, state->owner, 0, 0, NULL, GFP_NOFS);
+	opendata = nfs4_opendata_alloc(ctx->path.dentry, state->owner, 0, 0, NULL, GFP_NOFS);
 	if (opendata == NULL)
 		return ERR_PTR(-ENOMEM);
 	opendata->state = state;
@@ -1706,7 +1708,7 @@ static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata, struct
 /*
  * Returns a referenced nfs4_state
  */
-static int _nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res)
+static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res)
 {
 	struct nfs4_state_owner  *sp;
 	struct nfs4_state     *state = NULL;
@@ -1723,15 +1725,15 @@ static int _nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, in
 	status = nfs4_recover_expired_lease(server);
 	if (status != 0)
 		goto err_put_state_owner;
-	if (path->dentry->d_inode != NULL)
-		nfs4_return_incompatible_delegation(path->dentry->d_inode, fmode);
+	if (dentry->d_inode != NULL)
+		nfs4_return_incompatible_delegation(dentry->d_inode, fmode);
 	status = -ENOMEM;
-	opendata = nfs4_opendata_alloc(path, sp, fmode, flags, sattr, GFP_KERNEL);
+	opendata = nfs4_opendata_alloc(dentry, sp, fmode, flags, sattr, GFP_KERNEL);
 	if (opendata == NULL)
 		goto err_put_state_owner;
 
-	if (path->dentry->d_inode != NULL)
-		opendata->state = nfs4_get_open_state(path->dentry->d_inode, sp);
+	if (dentry->d_inode != NULL)
+		opendata->state = nfs4_get_open_state(dentry->d_inode, sp);
 
 	status = _nfs4_proc_open(opendata);
 	if (status != 0)
@@ -1769,14 +1771,14 @@ out_err:
 }
 
 
-static struct nfs4_state *nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred)
+static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred)
 {
 	struct nfs4_exception exception = { };
 	struct nfs4_state *res;
 	int status;
 
 	do {
-		status = _nfs4_do_open(dir, path, fmode, flags, sattr, cred, &res);
+		status = _nfs4_do_open(dir, dentry, fmode, flags, sattr, cred, &res);
 		if (status == 0)
 			break;
 		/* NOTE: BAD_SEQID means the server and client disagree about the
@@ -2079,7 +2081,7 @@ nfs4_atomic_open(struct inode *dir, struct nfs_open_context *ctx, int open_flags
 	struct nfs4_state *state;
 
 	/* Protect against concurrent sillydeletes */
-	state = nfs4_do_open(dir, &ctx->path, ctx->mode, open_flags, attr, ctx->cred);
+	state = nfs4_do_open(dir, ctx->path.dentry, ctx->mode, open_flags, attr, ctx->cred);
 	if (IS_ERR(state))
 		return ERR_CAST(state);
 	ctx->state = state;
@@ -2615,10 +2617,7 @@ static int
 nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
                  int flags, struct nfs_open_context *ctx)
 {
-	struct path my_path = {
-		.dentry = dentry,
-	};
-	struct path *path = &my_path;
+	struct dentry *de = dentry;
 	struct nfs4_state *state;
 	struct rpc_cred *cred = NULL;
 	fmode_t fmode = 0;
@@ -2626,11 +2625,11 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
 
 	if (ctx != NULL) {
 		cred = ctx->cred;
-		path = &ctx->path;
+		de = ctx->path.dentry;
 		fmode = ctx->mode;
 	}
 	sattr->ia_mode &= ~current_umask();
-	state = nfs4_do_open(dir, path, fmode, flags, sattr, cred);
+	state = nfs4_do_open(dir, de, fmode, flags, sattr, cred);
 	d_drop(dentry);
 	if (IS_ERR(state)) {
 		status = PTR_ERR(state);
-- 
cgit v1.2.3


From 3d4ff43d895c50319af45eb4bf04a4618eccdf76 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Wed, 22 Jun 2011 18:40:12 -0400
Subject: nfs_open_context doesn't need struct path either

just dentry, please...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfs/dir.c      | 14 +++++---------
 fs/nfs/direct.c   |  4 ++--
 fs/nfs/inode.c    | 20 +++++++++++---------
 fs/nfs/nfs4proc.c | 10 +++++-----
 fs/nfs/pagelist.c |  4 ++--
 fs/nfs/read.c     |  8 ++++----
 fs/nfs/write.c    | 22 +++++++++++-----------
 7 files changed, 40 insertions(+), 42 deletions(-)

(limited to 'fs')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index a86acd6d0cb4..a0693f3f11c0 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1345,10 +1345,6 @@ static int is_atomic_open(struct nameidata *nd)
 
 static struct nfs_open_context *nameidata_to_nfs_open_context(struct dentry *dentry, struct nameidata *nd)
 {
-	struct path path = {
-		.mnt = nd->path.mnt,
-		.dentry = dentry,
-	};
 	struct nfs_open_context *ctx;
 	struct rpc_cred *cred;
 	fmode_t fmode = nd->intent.open.flags & (FMODE_READ | FMODE_WRITE | FMODE_EXEC);
@@ -1356,7 +1352,7 @@ static struct nfs_open_context *nameidata_to_nfs_open_context(struct dentry *den
 	cred = rpc_lookup_cred();
 	if (IS_ERR(cred))
 		return ERR_CAST(cred);
-	ctx = alloc_nfs_open_context(&path, cred, fmode);
+	ctx = alloc_nfs_open_context(dentry, cred, fmode);
 	put_rpccred(cred);
 	if (ctx == NULL)
 		return ERR_PTR(-ENOMEM);
@@ -1376,13 +1372,13 @@ static int nfs_intent_set_file(struct nameidata *nd, struct nfs_open_context *ct
 
 	/* If the open_intent is for execute, we have an extra check to make */
 	if (ctx->mode & FMODE_EXEC) {
-		ret = nfs_may_open(ctx->path.dentry->d_inode,
+		ret = nfs_may_open(ctx->dentry->d_inode,
 				ctx->cred,
 				nd->intent.open.flags);
 		if (ret < 0)
 			goto out;
 	}
-	filp = lookup_instantiate_filp(nd, ctx->path.dentry, do_open);
+	filp = lookup_instantiate_filp(nd, ctx->dentry, do_open);
 	if (IS_ERR(filp))
 		ret = PTR_ERR(filp);
 	else
@@ -1463,8 +1459,8 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
 	res = d_add_unique(dentry, inode);
 	nfs_unblock_sillyrename(dentry->d_parent);
 	if (res != NULL) {
-		dput(ctx->path.dentry);
-		ctx->path.dentry = dget(res);
+		dput(ctx->dentry);
+		ctx->dentry = dget(res);
 		dentry = res;
 	}
 	err = nfs_intent_set_file(nd, ctx);
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 8eea25366717..b35d25b98da6 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -284,7 +284,7 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq,
 						loff_t pos)
 {
 	struct nfs_open_context *ctx = dreq->ctx;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 	unsigned long user_addr = (unsigned long)iov->iov_base;
 	size_t count = iov->iov_len;
 	size_t rsize = NFS_SERVER(inode)->rsize;
@@ -715,7 +715,7 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq,
 						 loff_t pos, int sync)
 {
 	struct nfs_open_context *ctx = dreq->ctx;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 	unsigned long user_addr = (unsigned long)iov->iov_base;
 	size_t count = iov->iov_len;
 	struct rpc_task *task;
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 6f4850deb272..fe1203797b2b 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -567,7 +567,7 @@ static struct nfs_lock_context *__nfs_find_lock_context(struct nfs_open_context
 struct nfs_lock_context *nfs_get_lock_context(struct nfs_open_context *ctx)
 {
 	struct nfs_lock_context *res, *new = NULL;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 
 	spin_lock(&inode->i_lock);
 	res = __nfs_find_lock_context(ctx);
@@ -594,7 +594,7 @@ struct nfs_lock_context *nfs_get_lock_context(struct nfs_open_context *ctx)
 void nfs_put_lock_context(struct nfs_lock_context *l_ctx)
 {
 	struct nfs_open_context *ctx = l_ctx->open_context;
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
 
 	if (!atomic_dec_and_lock(&l_ctx->count, &inode->i_lock))
 		return;
@@ -620,7 +620,7 @@ void nfs_close_context(struct nfs_open_context *ctx, int is_sync)
 		return;
 	if (!is_sync)
 		return;
-	inode = ctx->path.dentry->d_inode;
+	inode = ctx->dentry->d_inode;
 	if (!list_empty(&NFS_I(inode)->open_files))
 		return;
 	server = NFS_SERVER(inode);
@@ -629,14 +629,14 @@ void nfs_close_context(struct nfs_open_context *ctx, int is_sync)
 	nfs_revalidate_inode(server, inode);
 }
 
-struct nfs_open_context *alloc_nfs_open_context(struct path *path, struct rpc_cred *cred, fmode_t f_mode)
+struct nfs_open_context *alloc_nfs_open_context(struct dentry *dentry, struct rpc_cred *cred, fmode_t f_mode)
 {
 	struct nfs_open_context *ctx;
 
 	ctx = kmalloc(sizeof(*ctx), GFP_KERNEL);
 	if (ctx != NULL) {
-		ctx->path = *path;
-		path_get(&ctx->path);
+		nfs_sb_active(dentry->d_sb);
+		ctx->dentry = dget(dentry);
 		ctx->cred = get_rpccred(cred);
 		ctx->state = NULL;
 		ctx->mode = f_mode;
@@ -658,7 +658,8 @@ struct nfs_open_context *get_nfs_open_context(struct nfs_open_context *ctx)
 
 static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync)
 {
-	struct inode *inode = ctx->path.dentry->d_inode;
+	struct inode *inode = ctx->dentry->d_inode;
+	struct super_block *sb = ctx->dentry->d_sb;
 
 	if (!list_empty(&ctx->list)) {
 		if (!atomic_dec_and_lock(&ctx->lock_context.count, &inode->i_lock))
@@ -671,7 +672,8 @@ static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync)
 		NFS_PROTO(inode)->close_context(ctx, is_sync);
 	if (ctx->cred != NULL)
 		put_rpccred(ctx->cred);
-	path_put(&ctx->path);
+	dput(ctx->dentry);
+	nfs_sb_deactive(sb);
 	kfree(ctx);
 }
 
@@ -741,7 +743,7 @@ int nfs_open(struct inode *inode, struct file *filp)
 	cred = rpc_lookup_cred();
 	if (IS_ERR(cred))
 		return PTR_ERR(cred);
-	ctx = alloc_nfs_open_context(&filp->f_path, cred, filp->f_mode);
+	ctx = alloc_nfs_open_context(filp->f_path.dentry, cred, filp->f_mode);
 	put_rpccred(cred);
 	if (ctx == NULL)
 		return -ENOMEM;
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index aef22fa0a1c0..26bece8f3083 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1132,7 +1132,7 @@ static struct nfs4_opendata *nfs4_open_recoverdata_alloc(struct nfs_open_context
 {
 	struct nfs4_opendata *opendata;
 
-	opendata = nfs4_opendata_alloc(ctx->path.dentry, state->owner, 0, 0, NULL, GFP_NOFS);
+	opendata = nfs4_opendata_alloc(ctx->dentry, state->owner, 0, 0, NULL, GFP_NOFS);
 	if (opendata == NULL)
 		return ERR_PTR(-ENOMEM);
 	opendata->state = state;
@@ -1650,7 +1650,7 @@ static int _nfs4_open_expired(struct nfs_open_context *ctx, struct nfs4_state *s
 		return PTR_ERR(opendata);
 	ret = nfs4_open_recover(opendata, state);
 	if (ret == -ESTALE)
-		d_drop(ctx->path.dentry);
+		d_drop(ctx->dentry);
 	nfs4_opendata_put(opendata);
 	return ret;
 }
@@ -2081,7 +2081,7 @@ nfs4_atomic_open(struct inode *dir, struct nfs_open_context *ctx, int open_flags
 	struct nfs4_state *state;
 
 	/* Protect against concurrent sillydeletes */
-	state = nfs4_do_open(dir, ctx->path.dentry, ctx->mode, open_flags, attr, ctx->cred);
+	state = nfs4_do_open(dir, ctx->dentry, ctx->mode, open_flags, attr, ctx->cred);
 	if (IS_ERR(state))
 		return ERR_CAST(state);
 	ctx->state = state;
@@ -2625,7 +2625,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr,
 
 	if (ctx != NULL) {
 		cred = ctx->cred;
-		de = ctx->path.dentry;
+		de = ctx->dentry;
 		fmode = ctx->mode;
 	}
 	sattr->ia_mode &= ~current_umask();
@@ -4292,7 +4292,7 @@ static void nfs4_lock_done(struct rpc_task *task, void *calldata)
 		memcpy(data->lsp->ls_stateid.data, data->res.stateid.data,
 					sizeof(data->lsp->ls_stateid.data));
 		data->lsp->ls_flags |= NFS_LOCK_INITIALIZED;
-		renew_lease(NFS_SERVER(data->ctx->path.dentry->d_inode), data->timestamp);
+		renew_lease(NFS_SERVER(data->ctx->dentry->d_inode), data->timestamp);
 	}
 out:
 	dprintk("%s: done, ret = %d!\n", __func__, data->rpc_status);
diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index 009855716286..18449f43c568 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -114,7 +114,7 @@ int nfs_set_page_tag_locked(struct nfs_page *req)
 	if (!nfs_lock_request_dontget(req))
 		return 0;
 	if (test_bit(PG_MAPPED, &req->wb_flags))
-		radix_tree_tag_set(&NFS_I(req->wb_context->path.dentry->d_inode)->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED);
+		radix_tree_tag_set(&NFS_I(req->wb_context->dentry->d_inode)->nfs_page_tree, req->wb_index, NFS_PAGE_TAG_LOCKED);
 	return 1;
 }
 
@@ -124,7 +124,7 @@ int nfs_set_page_tag_locked(struct nfs_page *req)
 void nfs_clear_page_tag_locked(struct nfs_page *req)
 {
 	if (test_bit(PG_MAPPED, &req->wb_flags)) {
-		struct inode *inode = req->wb_context->path.dentry->d_inode;
+		struct inode *inode = req->wb_context->dentry->d_inode;
 		struct nfs_inode *nfsi = NFS_I(inode);
 
 		spin_lock(&inode->i_lock);
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index 20a7f952e244..a68679f538fc 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -144,7 +144,7 @@ int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode,
 
 static void nfs_readpage_release(struct nfs_page *req)
 {
-	struct inode *d_inode = req->wb_context->path.dentry->d_inode;
+	struct inode *d_inode = req->wb_context->dentry->d_inode;
 
 	if (PageUptodate(req->wb_page))
 		nfs_readpage_to_fscache(d_inode, req->wb_page, 0);
@@ -152,8 +152,8 @@ static void nfs_readpage_release(struct nfs_page *req)
 	unlock_page(req->wb_page);
 
 	dprintk("NFS: read done (%s/%Ld %d@%Ld)\n",
-			req->wb_context->path.dentry->d_inode->i_sb->s_id,
-			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
+			req->wb_context->dentry->d_inode->i_sb->s_id,
+			(long long)NFS_FILEID(req->wb_context->dentry->d_inode),
 			req->wb_bytes,
 			(long long)req_offset(req));
 	nfs_release_request(req);
@@ -207,7 +207,7 @@ static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data,
 		unsigned int count, unsigned int offset,
 		struct pnfs_layout_segment *lseg)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 
 	data->req	  = req;
 	data->inode	  = inode;
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 727168059684..08579312c57b 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -409,7 +409,7 @@ out:
  */
 static void nfs_inode_remove_request(struct nfs_page *req)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	BUG_ON (!NFS_WBACK_BUSY(req));
@@ -438,7 +438,7 @@ nfs_mark_request_dirty(struct nfs_page *req)
 static void
 nfs_mark_request_commit(struct nfs_page *req, struct pnfs_layout_segment *lseg)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	spin_lock(&inode->i_lock);
@@ -852,13 +852,13 @@ static int nfs_write_rpcsetup(struct nfs_page *req,
 		struct pnfs_layout_segment *lseg,
 		int how)
 {
-	struct inode *inode = req->wb_context->path.dentry->d_inode;
+	struct inode *inode = req->wb_context->dentry->d_inode;
 
 	/* Set up the RPC argument and reply structs
 	 * NB: take care not to mess about with data->commit et al. */
 
 	data->req = req;
-	data->inode = inode = req->wb_context->path.dentry->d_inode;
+	data->inode = inode = req->wb_context->dentry->d_inode;
 	data->cred = req->wb_context->cred;
 	data->lseg = get_lseg(lseg);
 
@@ -1053,9 +1053,9 @@ static void nfs_writeback_done_partial(struct rpc_task *task, void *calldata)
 
 	dprintk("NFS: %5u write(%s/%lld %d@%lld)",
 		task->tk_pid,
-		data->req->wb_context->path.dentry->d_inode->i_sb->s_id,
+		data->req->wb_context->dentry->d_inode->i_sb->s_id,
 		(long long)
-		  NFS_FILEID(data->req->wb_context->path.dentry->d_inode),
+		  NFS_FILEID(data->req->wb_context->dentry->d_inode),
 		data->req->wb_bytes, (long long)req_offset(data->req));
 
 	nfs_writeback_done(task, data);
@@ -1148,8 +1148,8 @@ static void nfs_writeback_release_full(void *calldata)
 
 		dprintk("NFS: %5u write (%s/%lld %d@%lld)",
 			data->task.tk_pid,
-			req->wb_context->path.dentry->d_inode->i_sb->s_id,
-			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
+			req->wb_context->dentry->d_inode->i_sb->s_id,
+			(long long)NFS_FILEID(req->wb_context->dentry->d_inode),
 			req->wb_bytes,
 			(long long)req_offset(req));
 
@@ -1347,7 +1347,7 @@ void nfs_init_commit(struct nfs_write_data *data,
 			    struct pnfs_layout_segment *lseg)
 {
 	struct nfs_page *first = nfs_list_entry(head->next);
-	struct inode *inode = first->wb_context->path.dentry->d_inode;
+	struct inode *inode = first->wb_context->dentry->d_inode;
 
 	/* Set up the RPC argument and reply structs
 	 * NB: take care not to mess about with data->commit et al. */
@@ -1435,8 +1435,8 @@ void nfs_commit_release_pages(struct nfs_write_data *data)
 		nfs_clear_request_commit(req);
 
 		dprintk("NFS:       commit (%s/%lld %d@%lld)",
-			req->wb_context->path.dentry->d_inode->i_sb->s_id,
-			(long long)NFS_FILEID(req->wb_context->path.dentry->d_inode),
+			req->wb_context->dentry->d_sb->s_id,
+			(long long)NFS_FILEID(req->wb_context->dentry->d_inode),
 			req->wb_bytes,
 			(long long)req_offset(req));
 		if (status < 0) {
-- 
cgit v1.2.3


From 511415980ace0ceecd71088dbe1c7ce4ca7c79fe Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Wed, 22 Jun 2011 18:47:28 -0400
Subject: nameidata_to_nfs_open_context() doesn't need nameidata, actually...

just open flags; switched to passing just those and
renamed to create_nfs_open_context()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfs/dir.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

(limited to 'fs')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index a0693f3f11c0..cc613c354bb8 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1343,11 +1343,11 @@ static int is_atomic_open(struct nameidata *nd)
 	return 1;
 }
 
-static struct nfs_open_context *nameidata_to_nfs_open_context(struct dentry *dentry, struct nameidata *nd)
+static struct nfs_open_context *create_nfs_open_context(struct dentry *dentry, int open_flags)
 {
 	struct nfs_open_context *ctx;
 	struct rpc_cred *cred;
-	fmode_t fmode = nd->intent.open.flags & (FMODE_READ | FMODE_WRITE | FMODE_EXEC);
+	fmode_t fmode = open_flags & (FMODE_READ | FMODE_WRITE | FMODE_EXEC);
 
 	cred = rpc_lookup_cred();
 	if (IS_ERR(cred))
@@ -1416,12 +1416,13 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
 		goto out;
 	}
 
-	ctx = nameidata_to_nfs_open_context(dentry, nd);
+	open_flags = nd->intent.open.flags;
+
+	ctx = create_nfs_open_context(dentry, open_flags);
 	res = ERR_CAST(ctx);
 	if (IS_ERR(ctx))
 		goto out;
 
-	open_flags = nd->intent.open.flags;
 	if (nd->flags & LOOKUP_CREATE) {
 		attr.ia_mode = nd->intent.open.create_mode;
 		attr.ia_valid = ATTR_MODE;
@@ -1513,7 +1514,7 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd)
 	/* We can't create new files, or truncate existing ones here */
 	openflags &= ~(O_CREAT|O_EXCL|O_TRUNC);
 
-	ctx = nameidata_to_nfs_open_context(dentry, nd);
+	ctx = create_nfs_open_context(dentry, openflags);
 	ret = PTR_ERR(ctx);
 	if (IS_ERR(ctx))
 		goto out;
@@ -1577,7 +1578,7 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	if ((nd->flags & LOOKUP_CREATE) != 0) {
 		open_flags = nd->intent.open.flags;
 
-		ctx = nameidata_to_nfs_open_context(dentry, nd);
+		ctx = create_nfs_open_context(dentry, open_flags);
 		error = PTR_ERR(ctx);
 		if (IS_ERR(ctx))
 			goto out_err_drop;
-- 
cgit v1.2.3


From f7c85868fcacc331dd3454a4f08f006d7942521f Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Wed, 22 Jun 2011 18:53:18 -0400
Subject: fix mknod() on nfs4 (hopefully)

a) check the right flags in ->create() (LOOKUP_OPEN, not LOOKUP_CREATE)
b) default (!LOOKUP_OPEN) open_flags is O_CREAT|O_EXCL|FMODE_READ, not 0
c) lookup_instantiate_filp() should be done only with LOOKUP_OPEN;
otherwise we need to issue CLOSE, lest we leak stateid on server.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfs/dir.c | 24 ++++++++++++------------
 1 file changed, 12 insertions(+), 12 deletions(-)

(limited to 'fs')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index cc613c354bb8..1f4625749038 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1567,7 +1567,7 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	struct nfs_open_context *ctx = NULL;
 	struct iattr attr;
 	int error;
-	int open_flags = 0;
+	int open_flags = O_CREAT|O_EXCL|FMODE_READ;
 
 	dfprintk(VFS, "NFS: create(%s/%ld), %s\n",
 			dir->i_sb->s_id, dir->i_ino, dentry->d_name.name);
@@ -1575,27 +1575,27 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	attr.ia_mode = mode;
 	attr.ia_valid = ATTR_MODE;
 
-	if ((nd->flags & LOOKUP_CREATE) != 0) {
+	if (nd && (nd->flags & LOOKUP_OPEN) != 0)
 		open_flags = nd->intent.open.flags;
 
-		ctx = create_nfs_open_context(dentry, open_flags);
-		error = PTR_ERR(ctx);
-		if (IS_ERR(ctx))
-			goto out_err_drop;
-	}
+	ctx = create_nfs_open_context(dentry, open_flags);
+	error = PTR_ERR(ctx);
+	if (IS_ERR(ctx))
+		goto out_err_drop;
 
 	error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, ctx);
 	if (error != 0)
 		goto out_put_ctx;
-	if (ctx != NULL) {
+	if (nd && (nd->flags & LOOKUP_OPEN) != 0) {
 		error = nfs_intent_set_file(nd, ctx);
 		if (error < 0)
 			goto out_err;
+	} else {
+		put_nfs_open_context(ctx);
 	}
 	return 0;
 out_put_ctx:
-	if (ctx != NULL)
-		put_nfs_open_context(ctx);
+	put_nfs_open_context(ctx);
 out_err_drop:
 	d_drop(dentry);
 out_err:
@@ -1657,7 +1657,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode,
 {
 	struct iattr attr;
 	int error;
-	int open_flags = 0;
+	int open_flags = O_CREAT|O_EXCL|FMODE_READ;
 
 	dfprintk(VFS, "NFS: create(%s/%ld), %s\n",
 			dir->i_sb->s_id, dir->i_ino, dentry->d_name.name);
@@ -1665,7 +1665,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	attr.ia_mode = mode;
 	attr.ia_valid = ATTR_MODE;
 
-	if ((nd->flags & LOOKUP_CREATE) != 0)
+	if (nd && (nd->flags & LOOKUP_OPEN) != 0)
 		open_flags = nd->intent.open.flags;
 
 	error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, NULL);
-- 
cgit v1.2.3


From 554a8b9f54cd7ca2b89f5dc227df08be082fae0d Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 23 Jun 2011 12:35:50 -0400
Subject: Don't pass nameidata when calling vfs_create() from mknod()

All instances can cope with that now (and ceph one actually
starts working properly).

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 42ccb97cc260..94fd0fa2d647 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -2411,7 +2411,7 @@ SYSCALL_DEFINE4(mknodat, int, dfd, const char __user *, filename, int, mode,
 		goto out_drop_write;
 	switch (mode & S_IFMT) {
 		case 0: case S_IFREG:
-			error = vfs_create(nd.path.dentry->d_inode,dentry,mode,&nd);
+			error = vfs_create(nd.path.dentry->d_inode,dentry,mode,NULL);
 			break;
 		case S_IFCHR: case S_IFBLK:
 			error = vfs_mknod(nd.path.dentry->d_inode,dentry,mode,
-- 
cgit v1.2.3


From 8a5e929dd2e05ab4d3d89f58c5e8fca596af8f3a Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 19:15:54 -0400
Subject: don't transliterate lower bits of ->intent.open.flags to FMODE_...

->create() instances are much happier that way...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/9p/vfs_inode.c      |  2 +-
 fs/9p/vfs_inode_dotl.c |  2 +-
 fs/ceph/file.c         |  2 +-
 fs/fuse/dir.c          |  2 +-
 fs/namei.c             | 21 ++-------------------
 fs/nfs/dir.c           | 24 +++++++++++++++++-------
 6 files changed, 23 insertions(+), 30 deletions(-)

(limited to 'fs')

diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c
index 7f6c67703195..47f71eb66b32 100644
--- a/fs/9p/vfs_inode.c
+++ b/fs/9p/vfs_inode.c
@@ -634,7 +634,7 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	v9ses = v9fs_inode2v9ses(dir);
 	perm = unixmode2p9mode(v9ses, mode);
 	if (nd && nd->flags & LOOKUP_OPEN)
-		flags = nd->intent.open.flags - 1;
+		flags = nd->intent.open.flags;
 	else
 		flags = O_RDWR;
 
diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c
index 691c78f58bef..d148e69f21b5 100644
--- a/fs/9p/vfs_inode_dotl.c
+++ b/fs/9p/vfs_inode_dotl.c
@@ -174,7 +174,7 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode,
 
 	v9ses = v9fs_inode2v9ses(dir);
 	if (nd && nd->flags & LOOKUP_OPEN)
-		flags = nd->intent.open.flags - 1;
+		flags = nd->intent.open.flags;
 	else {
 		/*
 		 * create call without LOOKUP_OPEN is due
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 4698a5c553dc..0a292459c895 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -226,7 +226,7 @@ struct dentry *ceph_lookup_open(struct inode *dir, struct dentry *dentry,
 	struct inode *parent_inode = get_dentry_parent_inode(file->f_dentry);
 	struct ceph_mds_request *req;
 	int err;
-	int flags = nd->intent.open.flags - 1;  /* silly vfs! */
+	int flags = nd->intent.open.flags;
 
 	dout("ceph_lookup_open dentry %p '%.*s' flags %d mode 0%o\n",
 	     dentry, dentry->d_name.len, dentry->d_name.name, flags, mode);
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index e2b14001cea5..47559dd33193 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -382,7 +382,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode,
 	struct fuse_entry_out outentry;
 	struct fuse_file *ff;
 	struct file *file;
-	int flags = nd->intent.open.flags - 1;
+	int flags = nd->intent.open.flags;
 
 	if (fc->no_create)
 		return -ENOSYS;
diff --git a/fs/namei.c b/fs/namei.c
index 94fd0fa2d647..5e65f67ee926 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1965,27 +1965,10 @@ static int handle_truncate(struct file *filp)
 	return error;
 }
 
-/*
- * Note that while the flag value (low two bits) for sys_open means:
- *	00 - read-only
- *	01 - write-only
- *	10 - read-write
- *	11 - special
- * it is changed into
- *	00 - no permissions needed
- *	01 - read-permission
- *	10 - write-permission
- *	11 - read-write
- * for the internal routines (ie open_namei()/follow_link() etc)
- * This is more logical, and also allows the 00 "no perm needed"
- * to be used for symlinks (where the permissions are checked
- * later).
- *
-*/
 static inline int open_to_namei_flags(int flag)
 {
-	if ((flag+1) & O_ACCMODE)
-		flag++;
+	if ((flag & O_ACCMODE) == 3)
+		flag--;
 	return flag;
 }
 
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 1f4625749038..b5f63a50fa7f 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1338,16 +1338,26 @@ static int is_atomic_open(struct nameidata *nd)
 		return 0;
 	/* Are we trying to write to a read only partition? */
 	if (__mnt_is_readonly(nd->path.mnt) &&
-	    (nd->intent.open.flags & (O_CREAT|O_TRUNC|FMODE_WRITE)))
+	    (nd->intent.open.flags & (O_CREAT|O_TRUNC|O_ACCMODE)))
 		return 0;
 	return 1;
 }
 
+static fmode_t flags_to_mode(int flags)
+{
+	fmode_t res = (__force fmode_t)flags & FMODE_EXEC;
+	if ((flags & O_ACCMODE) != O_WRONLY)
+		res |= FMODE_READ;
+	if ((flags & O_ACCMODE) != O_RDONLY)
+		res |= FMODE_WRITE;
+	return res;
+}
+
 static struct nfs_open_context *create_nfs_open_context(struct dentry *dentry, int open_flags)
 {
 	struct nfs_open_context *ctx;
 	struct rpc_cred *cred;
-	fmode_t fmode = open_flags & (FMODE_READ | FMODE_WRITE | FMODE_EXEC);
+	fmode_t fmode = flags_to_mode(open_flags);
 
 	cred = rpc_lookup_cred();
 	if (IS_ERR(cred))
@@ -1567,7 +1577,7 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	struct nfs_open_context *ctx = NULL;
 	struct iattr attr;
 	int error;
-	int open_flags = O_CREAT|O_EXCL|FMODE_READ;
+	int open_flags = O_CREAT|O_EXCL;
 
 	dfprintk(VFS, "NFS: create(%s/%ld), %s\n",
 			dir->i_sb->s_id, dir->i_ino, dentry->d_name.name);
@@ -1657,7 +1667,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode,
 {
 	struct iattr attr;
 	int error;
-	int open_flags = O_CREAT|O_EXCL|FMODE_READ;
+	int open_flags = O_CREAT|O_EXCL;
 
 	dfprintk(VFS, "NFS: create(%s/%ld), %s\n",
 			dir->i_sb->s_id, dir->i_ino, dentry->d_name.name);
@@ -2256,11 +2266,11 @@ static int nfs_open_permission_mask(int openflags)
 {
 	int mask = 0;
 
-	if (openflags & FMODE_READ)
+	if ((openflags & O_ACCMODE) != O_WRONLY)
 		mask |= MAY_READ;
-	if (openflags & FMODE_WRITE)
+	if ((openflags & O_ACCMODE) != O_RDONLY)
 		mask |= MAY_WRITE;
-	if (openflags & FMODE_EXEC)
+	if (openflags & __FMODE_EXEC)
 		mask |= MAY_EXEC;
 	return mask;
 }
-- 
cgit v1.2.3


From bf6c7f6c7bd0ea779757d35b5fdc9f9157f056b3 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:08:31 -0400
Subject: don't pass nameidata to vfs_create() from ecryptfs_create()

Instead of playing with removal of LOOKUP_OPEN, mangling (and
restoring) nd->path, just pass NULL to vfs_create().  The whole
point of what's being done there is to suppress any attempts
to open file by underlying fs, which is what nd == NULL indicates.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ecryptfs/inode.c | 33 +++++----------------------------
 1 file changed, 5 insertions(+), 28 deletions(-)

(limited to 'fs')

diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 799e01055f22..340c657a108c 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -147,7 +147,6 @@ static int ecryptfs_interpose(struct dentry *lower_dentry,
  * @lower_dir_inode: inode of the parent in the lower fs of the new file
  * @dentry: New file's dentry
  * @mode: The mode of the new file
- * @nd: nameidata of ecryptfs' parent's dentry & vfsmount
  *
  * Creates the file in the lower file system.
  *
@@ -155,31 +154,10 @@ static int ecryptfs_interpose(struct dentry *lower_dentry,
  */
 static int
 ecryptfs_create_underlying_file(struct inode *lower_dir_inode,
-				struct dentry *dentry, int mode,
-				struct nameidata *nd)
+				struct dentry *dentry, int mode)
 {
 	struct dentry *lower_dentry = ecryptfs_dentry_to_lower(dentry);
-	struct vfsmount *lower_mnt = ecryptfs_dentry_to_lower_mnt(dentry);
-	struct dentry *dentry_save;
-	struct vfsmount *vfsmount_save;
-	unsigned int flags_save;
-	int rc;
-
-	if (nd) {
-		dentry_save = nd->path.dentry;
-		vfsmount_save = nd->path.mnt;
-		flags_save = nd->flags;
-		nd->path.dentry = lower_dentry;
-		nd->path.mnt = lower_mnt;
-		nd->flags &= ~LOOKUP_OPEN;
-	}
-	rc = vfs_create(lower_dir_inode, lower_dentry, mode, nd);
-	if (nd) {
-		nd->path.dentry = dentry_save;
-		nd->path.mnt = vfsmount_save;
-		nd->flags = flags_save;
-	}
-	return rc;
+	return vfs_create(lower_dir_inode, lower_dentry, mode, NULL);
 }
 
 /**
@@ -197,8 +175,7 @@ ecryptfs_create_underlying_file(struct inode *lower_dir_inode,
  */
 static int
 ecryptfs_do_create(struct inode *directory_inode,
-		   struct dentry *ecryptfs_dentry, int mode,
-		   struct nameidata *nd)
+		   struct dentry *ecryptfs_dentry, int mode)
 {
 	int rc;
 	struct dentry *lower_dentry;
@@ -213,7 +190,7 @@ ecryptfs_do_create(struct inode *directory_inode,
 		goto out;
 	}
 	rc = ecryptfs_create_underlying_file(lower_dir_dentry->d_inode,
-					     ecryptfs_dentry, mode, nd);
+					     ecryptfs_dentry, mode);
 	if (rc) {
 		printk(KERN_ERR "%s: Failure to create dentry in lower fs; "
 		       "rc = [%d]\n", __func__, rc);
@@ -294,7 +271,7 @@ ecryptfs_create(struct inode *directory_inode, struct dentry *ecryptfs_dentry,
 	int rc;
 
 	/* ecryptfs_do_create() calls ecryptfs_interpose() */
-	rc = ecryptfs_do_create(directory_inode, ecryptfs_dentry, mode, nd);
+	rc = ecryptfs_do_create(directory_inode, ecryptfs_dentry, mode);
 	if (unlikely(rc)) {
 		ecryptfs_printk(KERN_WARNING, "Failed to create file in"
 				"lower filesystem\n");
-- 
cgit v1.2.3


From dd7dd556e45133ef13f2c4bddc0e0b1ac23bc0e4 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:17:17 -0400
Subject: no need to check for LOOKUP_OPEN in ->create() instances

... it will be set in nd->flag for all cases with non-NULL nd
(i.e. when called from do_last()).

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/9p/vfs_inode.c      | 4 ++--
 fs/9p/vfs_inode_dotl.c | 2 +-
 fs/cifs/dir.c          | 6 +++---
 fs/fuse/dir.c          | 2 +-
 fs/nfs/dir.c           | 6 +++---
 5 files changed, 10 insertions(+), 10 deletions(-)

(limited to 'fs')

diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c
index 47f71eb66b32..7f9976a866e9 100644
--- a/fs/9p/vfs_inode.c
+++ b/fs/9p/vfs_inode.c
@@ -633,7 +633,7 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	fid = NULL;
 	v9ses = v9fs_inode2v9ses(dir);
 	perm = unixmode2p9mode(v9ses, mode);
-	if (nd && nd->flags & LOOKUP_OPEN)
+	if (nd)
 		flags = nd->intent.open.flags;
 	else
 		flags = O_RDWR;
@@ -649,7 +649,7 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 
 	v9fs_invalidate_inode_attr(dir);
 	/* if we are opening a file, assign the open fid to the file */
-	if (nd && nd->flags & LOOKUP_OPEN) {
+	if (nd) {
 		v9inode = V9FS_I(dentry->d_inode);
 		mutex_lock(&v9inode->v_mutex);
 		if (v9ses->cache && !v9inode->writeback_fid &&
diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c
index d148e69f21b5..32bbbe5aa689 100644
--- a/fs/9p/vfs_inode_dotl.c
+++ b/fs/9p/vfs_inode_dotl.c
@@ -173,7 +173,7 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode,
 	struct posix_acl *pacl = NULL, *dacl = NULL;
 
 	v9ses = v9fs_inode2v9ses(dir);
-	if (nd && nd->flags & LOOKUP_OPEN)
+	if (nd)
 		flags = nd->intent.open.flags;
 	else {
 		/*
diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c
index fa8c21d913bc..8766149f6300 100644
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -179,7 +179,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode,
 	if (oplockEnabled)
 		oplock = REQ_OPLOCK;
 
-	if (nd && (nd->flags & LOOKUP_OPEN))
+	if (nd)
 		oflags = nd->intent.open.file->f_flags;
 	else
 		oflags = O_RDONLY | O_CREAT;
@@ -214,7 +214,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode,
 		   which should be rare for path not covered on files) */
 	}
 
-	if (nd && (nd->flags & LOOKUP_OPEN)) {
+	if (nd) {
 		/* if the file is going to stay open, then we
 		   need to set the desired access properly */
 		desiredAccess = 0;
@@ -328,7 +328,7 @@ cifs_create_set_dentry:
 	else
 		cFYI(1, "Create worked, get_inode_info failed rc = %d", rc);
 
-	if (newinode && nd && (nd->flags & LOOKUP_OPEN)) {
+	if (newinode && nd) {
 		struct cifsFileInfo *pfile_info;
 		struct file *filp;
 
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 47559dd33193..02063dde2728 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -576,7 +576,7 @@ static int fuse_mknod(struct inode *dir, struct dentry *entry, int mode,
 static int fuse_create(struct inode *dir, struct dentry *entry, int mode,
 		       struct nameidata *nd)
 {
-	if (nd && (nd->flags & LOOKUP_OPEN)) {
+	if (nd) {
 		int err = fuse_create_open(dir, entry, mode, nd);
 		if (err != -ENOSYS)
 			return err;
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index b5f63a50fa7f..77ae95f15497 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -1585,7 +1585,7 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	attr.ia_mode = mode;
 	attr.ia_valid = ATTR_MODE;
 
-	if (nd && (nd->flags & LOOKUP_OPEN) != 0)
+	if (nd)
 		open_flags = nd->intent.open.flags;
 
 	ctx = create_nfs_open_context(dentry, open_flags);
@@ -1596,7 +1596,7 @@ static int nfs_open_create(struct inode *dir, struct dentry *dentry, int mode,
 	error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, ctx);
 	if (error != 0)
 		goto out_put_ctx;
-	if (nd && (nd->flags & LOOKUP_OPEN) != 0) {
+	if (nd) {
 		error = nfs_intent_set_file(nd, ctx);
 		if (error < 0)
 			goto out_err;
@@ -1675,7 +1675,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode,
 	attr.ia_mode = mode;
 	attr.ia_valid = ATTR_MODE;
 
-	if (nd && (nd->flags & LOOKUP_OPEN) != 0)
+	if (nd)
 		open_flags = nd->intent.open.flags;
 
 	error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, NULL);
-- 
cgit v1.2.3


From 407938e79edcadba1b5a17cf928584d8a191a8d7 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:37:18 -0400
Subject: LOOKUP_CREATE and LOOKUP_RENAME_TARGET can be set only on the last
 step

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/cifs/dir.c       | 6 ++----
 fs/fat/namei_vfat.c | 6 ++----
 fs/jfs/namei.c      | 6 ++----
 3 files changed, 6 insertions(+), 12 deletions(-)

(limited to 'fs')

diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c
index 8766149f6300..251c2ca569d3 100644
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -663,10 +663,8 @@ cifs_d_revalidate(struct dentry *direntry, struct nameidata *nd)
 	 * case sensitive name which is specified by user if this is
 	 * for creation.
 	 */
-	if (!(nd->flags & (LOOKUP_CONTINUE | LOOKUP_PARENT))) {
-		if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
-			return 0;
-	}
+	if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
+		return 0;
 
 	if (time_after(jiffies, direntry->d_time + HZ) || !lookupCacheEnabled)
 		return 0;
diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c
index 20b4ea53fdc4..bb3f29c3557b 100644
--- a/fs/fat/namei_vfat.c
+++ b/fs/fat/namei_vfat.c
@@ -82,10 +82,8 @@ static int vfat_revalidate_ci(struct dentry *dentry, struct nameidata *nd)
 	 * case sensitive name which is specified by user if this is
 	 * for creation.
 	 */
-	if (!(nd->flags & (LOOKUP_CONTINUE | LOOKUP_PARENT))) {
-		if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
-			return 0;
-	}
+	if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
+		return 0;
 
 	return vfat_revalidate_shortname(dentry);
 }
diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index eaaf2b511e89..7a2e8e5152fd 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -1624,10 +1624,8 @@ static int jfs_ci_revalidate(struct dentry *dentry, struct nameidata *nd)
 	 * case sensitive name which is specified by user if this is
 	 * for creation.
 	 */
-	if (!(nd->flags & (LOOKUP_CONTINUE | LOOKUP_PARENT))) {
-		if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
-			return 0;
-	}
+	if (nd->flags & (LOOKUP_CREATE | LOOKUP_RENAME_TARGET))
+		return 0;
 	return 1;
 }
 
-- 
cgit v1.2.3


From 5c0f360b083fb33d05d1bff4b138b82d715eb419 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:41:09 -0400
Subject: jfs_ci_revalidate() is safe from RCU mode

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/jfs/namei.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'fs')

diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index 7a2e8e5152fd..1da0dc799286 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -1597,8 +1597,6 @@ out:
 
 static int jfs_ci_revalidate(struct dentry *dentry, struct nameidata *nd)
 {
-	if (nd && nd->flags & LOOKUP_RCU)
-		return -ECHILD;
 	/*
 	 * This is not negative dentry. Always valid.
 	 *
-- 
cgit v1.2.3


From a127e0af59ab610e8b37ac2dfa4cdb2ec8c8f604 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:43:56 -0400
Subject: ceph: LOOKUP_OPEN is set only when it's the last component

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ceph/dir.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'fs')

diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index ef8f08c343e8..b6d436f3c1ae 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -566,7 +566,6 @@ static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry,
 	/* open (but not create!) intent? */
 	if (nd &&
 	    (nd->flags & LOOKUP_OPEN) &&
-	    (nd->flags & LOOKUP_CONTINUE) == 0 && /* only open last component */
 	    !(nd->intent.open.flags & O_CREAT)) {
 		int mode = nd->intent.open.create_mode & ~current->fs->umask;
 		return ceph_lookup_open(dir, dentry, nd, mode, 1);
-- 
cgit v1.2.3


From 4352780386139ff33d2203868b392e6535deff61 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:45:21 -0400
Subject: cifs_lookup(): LOOKUP_OPEN is set only on the last component

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/cifs/dir.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c
index 251c2ca569d3..14d602f178c2 100644
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -568,7 +568,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry,
 	 * reduction in network traffic in the other paths.
 	 */
 	if (pTcon->unix_ext) {
-		if (nd && !(nd->flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY)) &&
+		if (nd && !(nd->flags & LOOKUP_DIRECTORY) &&
 		     (nd->flags & LOOKUP_OPEN) && !pTcon->broken_posix_open &&
 		     (nd->intent.open.file->f_flags & O_CREAT)) {
 			rc = cifs_posix_open(full_path, &newInode,
-- 
cgit v1.2.3


From 8aeb376ca0fe61038166c3b8243c678addb80abf Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:48:43 -0400
Subject: nfs: LOOKUP_{OPEN,CREATE,EXCL} is set only on the last step

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfs/dir.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

(limited to 'fs')

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 77ae95f15497..8a45e6d1f6a4 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -997,14 +997,12 @@ static int nfs_check_verifier(struct inode *dir, struct dentry *dentry)
  * Return the intent data that applies to this particular path component
  *
  * Note that the current set of intents only apply to the very last
- * component of the path.
- * We check for this using LOOKUP_CONTINUE and LOOKUP_PARENT.
+ * component of the path and none of them is set before that last
+ * component.
  */
 static inline unsigned int nfs_lookup_check_intent(struct nameidata *nd,
 						unsigned int mask)
 {
-	if (nd->flags & (LOOKUP_CONTINUE|LOOKUP_PARENT))
-		return 0;
 	return nd->flags & mask;
 }
 
-- 
cgit v1.2.3


From 49084c3bb2055c401f3493c13edae14d49128ca0 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 25 Jun 2011 21:59:52 -0400
Subject: kill LOOKUP_CONTINUE

LOOKUP_PARENT is equivalent to it now

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 5e65f67ee926..f49d6abfa799 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -664,7 +664,7 @@ static int follow_automount(struct path *path, unsigned flags,
 	/* We don't want to mount if someone supplied AT_NO_AUTOMOUNT
 	 * and this is the terminal part of the path.
 	 */
-	if ((flags & LOOKUP_NO_AUTOMOUNT) && !(flags & LOOKUP_CONTINUE))
+	if ((flags & LOOKUP_NO_AUTOMOUNT) && !(flags & LOOKUP_PARENT))
 		return -EISDIR; /* we actually want to stop here */
 
 	/* We want to mount if someone is trying to open/create a file of any
@@ -676,7 +676,7 @@ static int follow_automount(struct path *path, unsigned flags,
 	 * appended a '/' to the name.
 	 */
 	if (!(flags & LOOKUP_FOLLOW) &&
-	    !(flags & (LOOKUP_CONTINUE | LOOKUP_DIRECTORY |
+	    !(flags & (LOOKUP_PARENT | LOOKUP_DIRECTORY |
 		       LOOKUP_OPEN | LOOKUP_CREATE)))
 		return -EISDIR;
 
@@ -695,7 +695,7 @@ static int follow_automount(struct path *path, unsigned flags,
 		 * the path being looked up; if it wasn't then the remainder of
 		 * the path is inaccessible and we should say so.
 		 */
-		if (PTR_ERR(mnt) == -EISDIR && (flags & LOOKUP_CONTINUE))
+		if (PTR_ERR(mnt) == -EISDIR && (flags & LOOKUP_PARENT))
 			return -EREMOTE;
 		return PTR_ERR(mnt);
 	}
@@ -1281,7 +1281,6 @@ static int link_path_walk(const char *name, struct nameidata *nd)
 {
 	struct path next;
 	int err;
-	unsigned int lookup_flags = nd->flags;
 	
 	while (*name=='/')
 		name++;
@@ -1295,8 +1294,6 @@ static int link_path_walk(const char *name, struct nameidata *nd)
 		unsigned int c;
 		int type;
 
-		nd->flags |= LOOKUP_CONTINUE;
-
 		err = may_lookup(nd);
  		if (err)
 			break;
@@ -1358,8 +1355,6 @@ static int link_path_walk(const char *name, struct nameidata *nd)
 		/* here ends the main loop */
 
 last_component:
-		/* Clear LOOKUP_CONTINUE iff it was previously unset */
-		nd->flags &= lookup_flags | ~LOOKUP_CONTINUE;
 		nd->last = this;
 		nd->last_type = type;
 		return 0;
-- 
cgit v1.2.3


From dae6ad8f37529963ae7df52baaccf056b38f210e Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 26 Jun 2011 11:50:15 -0400
Subject: new helpers: kern_path_create/user_path_create

combination of kern_path_parent() and lookup_create().  Does *not*
expose struct nameidata to caller.  Syscalls converted to that...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c              | 154 ++++++++++++++++++++++++------------------------
 fs/ocfs2/refcounttree.c |  49 ++++-----------
 2 files changed, 87 insertions(+), 116 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index f49d6abfa799..b292eb03d9d2 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -2311,6 +2311,35 @@ fail:
 }
 EXPORT_SYMBOL_GPL(lookup_create);
 
+struct dentry *kern_path_create(int dfd, const char *pathname, struct path *path, int is_dir)
+{
+	struct nameidata nd;
+	struct dentry *res;
+	int error = do_path_lookup(dfd, pathname, LOOKUP_PARENT, &nd);
+	if (error)
+		return ERR_PTR(error);
+	res = lookup_create(&nd, is_dir);
+	if (IS_ERR(res)) {
+		mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
+		path_put(&nd.path);
+	}
+	*path = nd.path;
+	return res;
+}
+EXPORT_SYMBOL(kern_path_create);
+
+struct dentry *user_path_create(int dfd, const char __user *pathname, struct path *path, int is_dir)
+{
+	char *tmp = getname(pathname);
+	struct dentry *res;
+	if (IS_ERR(tmp))
+		return ERR_CAST(tmp);
+	res = kern_path_create(dfd, tmp, path, is_dir);
+	putname(tmp);
+	return res;
+}
+EXPORT_SYMBOL(user_path_create);
+
 int vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev)
 {
 	int error = may_create(dir, dentry);
@@ -2359,54 +2388,46 @@ static int may_mknod(mode_t mode)
 SYSCALL_DEFINE4(mknodat, int, dfd, const char __user *, filename, int, mode,
 		unsigned, dev)
 {
-	int error;
-	char *tmp;
 	struct dentry *dentry;
-	struct nameidata nd;
+	struct path path;
+	int error;
 
 	if (S_ISDIR(mode))
 		return -EPERM;
 
-	error = user_path_parent(dfd, filename, &nd, &tmp);
-	if (error)
-		return error;
+	dentry = user_path_create(dfd, filename, &path, 0);
+	if (IS_ERR(dentry))
+		return PTR_ERR(dentry);
 
-	dentry = lookup_create(&nd, 0);
-	if (IS_ERR(dentry)) {
-		error = PTR_ERR(dentry);
-		goto out_unlock;
-	}
-	if (!IS_POSIXACL(nd.path.dentry->d_inode))
+	if (!IS_POSIXACL(path.dentry->d_inode))
 		mode &= ~current_umask();
 	error = may_mknod(mode);
 	if (error)
 		goto out_dput;
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_mknod(&nd.path, dentry, mode, dev);
+	error = security_path_mknod(&path, dentry, mode, dev);
 	if (error)
 		goto out_drop_write;
 	switch (mode & S_IFMT) {
 		case 0: case S_IFREG:
-			error = vfs_create(nd.path.dentry->d_inode,dentry,mode,NULL);
+			error = vfs_create(path.dentry->d_inode,dentry,mode,NULL);
 			break;
 		case S_IFCHR: case S_IFBLK:
-			error = vfs_mknod(nd.path.dentry->d_inode,dentry,mode,
+			error = vfs_mknod(path.dentry->d_inode,dentry,mode,
 					new_decode_dev(dev));
 			break;
 		case S_IFIFO: case S_IFSOCK:
-			error = vfs_mknod(nd.path.dentry->d_inode,dentry,mode,0);
+			error = vfs_mknod(path.dentry->d_inode,dentry,mode,0);
 			break;
 	}
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(path.mnt);
 out_dput:
 	dput(dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-	path_put(&nd.path);
-	putname(tmp);
+	mutex_unlock(&path.dentry->d_inode->i_mutex);
+	path_put(&path);
 
 	return error;
 }
@@ -2439,38 +2460,29 @@ int vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 
 SYSCALL_DEFINE3(mkdirat, int, dfd, const char __user *, pathname, int, mode)
 {
-	int error = 0;
-	char * tmp;
 	struct dentry *dentry;
-	struct nameidata nd;
-
-	error = user_path_parent(dfd, pathname, &nd, &tmp);
-	if (error)
-		goto out_err;
+	struct path path;
+	int error;
 
-	dentry = lookup_create(&nd, 1);
-	error = PTR_ERR(dentry);
+	dentry = user_path_create(dfd, pathname, &path, 1);
 	if (IS_ERR(dentry))
-		goto out_unlock;
+		return PTR_ERR(dentry);
 
-	if (!IS_POSIXACL(nd.path.dentry->d_inode))
+	if (!IS_POSIXACL(path.dentry->d_inode))
 		mode &= ~current_umask();
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_mkdir(&nd.path, dentry, mode);
+	error = security_path_mkdir(&path, dentry, mode);
 	if (error)
 		goto out_drop_write;
-	error = vfs_mkdir(nd.path.dentry->d_inode, dentry, mode);
+	error = vfs_mkdir(path.dentry->d_inode, dentry, mode);
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(path.mnt);
 out_dput:
 	dput(dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-	path_put(&nd.path);
-	putname(tmp);
-out_err:
+	mutex_unlock(&path.dentry->d_inode->i_mutex);
+	path_put(&path);
 	return error;
 }
 
@@ -2730,38 +2742,31 @@ SYSCALL_DEFINE3(symlinkat, const char __user *, oldname,
 {
 	int error;
 	char *from;
-	char *to;
 	struct dentry *dentry;
-	struct nameidata nd;
+	struct path path;
 
 	from = getname(oldname);
 	if (IS_ERR(from))
 		return PTR_ERR(from);
 
-	error = user_path_parent(newdfd, newname, &nd, &to);
-	if (error)
-		goto out_putname;
-
-	dentry = lookup_create(&nd, 0);
+	dentry = user_path_create(newdfd, newname, &path, 0);
 	error = PTR_ERR(dentry);
 	if (IS_ERR(dentry))
-		goto out_unlock;
+		goto out_putname;
 
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_symlink(&nd.path, dentry, from);
+	error = security_path_symlink(&path, dentry, from);
 	if (error)
 		goto out_drop_write;
-	error = vfs_symlink(nd.path.dentry->d_inode, dentry, from);
+	error = vfs_symlink(path.dentry->d_inode, dentry, from);
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(path.mnt);
 out_dput:
 	dput(dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-	path_put(&nd.path);
-	putname(to);
+	mutex_unlock(&path.dentry->d_inode->i_mutex);
+	path_put(&path);
 out_putname:
 	putname(from);
 	return error;
@@ -2826,11 +2831,9 @@ SYSCALL_DEFINE5(linkat, int, olddfd, const char __user *, oldname,
 		int, newdfd, const char __user *, newname, int, flags)
 {
 	struct dentry *new_dentry;
-	struct nameidata nd;
-	struct path old_path;
+	struct path old_path, new_path;
 	int how = 0;
 	int error;
-	char *to;
 
 	if ((flags & ~(AT_SYMLINK_FOLLOW | AT_EMPTY_PATH)) != 0)
 		return -EINVAL;
@@ -2852,32 +2855,27 @@ SYSCALL_DEFINE5(linkat, int, olddfd, const char __user *, oldname,
 	if (error)
 		return error;
 
-	error = user_path_parent(newdfd, newname, &nd, &to);
-	if (error)
-		goto out;
-	error = -EXDEV;
-	if (old_path.mnt != nd.path.mnt)
-		goto out_release;
-	new_dentry = lookup_create(&nd, 0);
+	new_dentry = user_path_create(newdfd, newname, &new_path, 0);
 	error = PTR_ERR(new_dentry);
 	if (IS_ERR(new_dentry))
-		goto out_unlock;
-	error = mnt_want_write(nd.path.mnt);
+		goto out;
+
+	error = -EXDEV;
+	if (old_path.mnt != new_path.mnt)
+		goto out_dput;
+	error = mnt_want_write(new_path.mnt);
 	if (error)
 		goto out_dput;
-	error = security_path_link(old_path.dentry, &nd.path, new_dentry);
+	error = security_path_link(old_path.dentry, &new_path, new_dentry);
 	if (error)
 		goto out_drop_write;
-	error = vfs_link(old_path.dentry, nd.path.dentry->d_inode, new_dentry);
+	error = vfs_link(old_path.dentry, new_path.dentry->d_inode, new_dentry);
 out_drop_write:
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(new_path.mnt);
 out_dput:
 	dput(new_dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-out_release:
-	path_put(&nd.path);
-	putname(to);
+	mutex_unlock(&new_path.dentry->d_inode->i_mutex);
+	path_put(&new_path);
 out:
 	path_put(&old_path);
 
diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index ebfd3825f12a..cf7823382664 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -4368,25 +4368,6 @@ static inline int ocfs2_may_create(struct inode *dir, struct dentry *child)
 	return inode_permission(dir, MAY_WRITE | MAY_EXEC);
 }
 
-/* copied from user_path_parent. */
-static int ocfs2_user_path_parent(const char __user *path,
-				  struct nameidata *nd, char **name)
-{
-	char *s = getname(path);
-	int error;
-
-	if (IS_ERR(s))
-		return PTR_ERR(s);
-
-	error = kern_path_parent(s, nd);
-	if (error)
-		putname(s);
-	else
-		*name = s;
-
-	return error;
-}
-
 /**
  * ocfs2_vfs_reflink - Create a reference-counted link
  *
@@ -4460,10 +4441,8 @@ int ocfs2_reflink_ioctl(struct inode *inode,
 			bool preserve)
 {
 	struct dentry *new_dentry;
-	struct nameidata nd;
-	struct path old_path;
+	struct path old_path, new_path;
 	int error;
-	char *to = NULL;
 
 	if (!ocfs2_refcount_tree(OCFS2_SB(inode->i_sb)))
 		return -EOPNOTSUPP;
@@ -4474,39 +4453,33 @@ int ocfs2_reflink_ioctl(struct inode *inode,
 		return error;
 	}
 
-	error = ocfs2_user_path_parent(newname, &nd, &to);
-	if (error) {
+	new_dentry = user_path_create(AT_FDCWD, newname, &new_path, 0);
+	error = PTR_ERR(new_dentry);
+	if (IS_ERR(new_dentry)) {
 		mlog_errno(error);
 		goto out;
 	}
 
 	error = -EXDEV;
-	if (old_path.mnt != nd.path.mnt)
-		goto out_release;
-	new_dentry = lookup_create(&nd, 0);
-	error = PTR_ERR(new_dentry);
-	if (IS_ERR(new_dentry)) {
+	if (old_path.mnt != new_path.mnt) {
 		mlog_errno(error);
-		goto out_unlock;
+		goto out_dput;
 	}
 
-	error = mnt_want_write(nd.path.mnt);
+	error = mnt_want_write(new_path.mnt);
 	if (error) {
 		mlog_errno(error);
 		goto out_dput;
 	}
 
 	error = ocfs2_vfs_reflink(old_path.dentry,
-				  nd.path.dentry->d_inode,
+				  new_path.dentry->d_inode,
 				  new_dentry, preserve);
-	mnt_drop_write(nd.path.mnt);
+	mnt_drop_write(new_path.mnt);
 out_dput:
 	dput(new_dentry);
-out_unlock:
-	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-out_release:
-	path_put(&nd.path);
-	putname(to);
+	mutex_unlock(&new_path.dentry->d_inode->i_mutex);
+	path_put(&new_path);
 out:
 	path_put(&old_path);
 
-- 
cgit v1.2.3


From ed75e95de574c99575e5f3e1d9ca59ea8c12a9cb Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 27 Jun 2011 16:53:43 -0400
Subject: kill lookup_create()

folded into the only caller (kern_path_create())

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 54 ++++++++++++++++++------------------------------------
 1 file changed, 18 insertions(+), 36 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index b292eb03d9d2..b45a039216c7 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -2258,35 +2258,29 @@ struct file *do_file_open_root(struct dentry *dentry, struct vfsmount *mnt,
 	return file;
 }
 
-/**
- * lookup_create - lookup a dentry, creating it if it doesn't exist
- * @nd: nameidata info
- * @is_dir: directory flag
- *
- * Simple function to lookup and return a dentry and create it
- * if it doesn't exist.  Is SMP-safe.
- *
- * Returns with nd->path.dentry->d_inode->i_mutex locked.
- */
-struct dentry *lookup_create(struct nameidata *nd, int is_dir)
+struct dentry *kern_path_create(int dfd, const char *pathname, struct path *path, int is_dir)
 {
 	struct dentry *dentry = ERR_PTR(-EEXIST);
+	struct nameidata nd;
+	int error = do_path_lookup(dfd, pathname, LOOKUP_PARENT, &nd);
+	if (error)
+		return ERR_PTR(error);
 
-	mutex_lock_nested(&nd->path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
 	/*
 	 * Yucky last component or no last component at all?
 	 * (foo/., foo/.., /////)
 	 */
-	if (nd->last_type != LAST_NORM)
-		goto fail;
-	nd->flags &= ~LOOKUP_PARENT;
-	nd->flags |= LOOKUP_CREATE | LOOKUP_EXCL;
-	nd->intent.open.flags = O_EXCL;
+	if (nd.last_type != LAST_NORM)
+		goto out;
+	nd.flags &= ~LOOKUP_PARENT;
+	nd.flags |= LOOKUP_CREATE | LOOKUP_EXCL;
+	nd.intent.open.flags = O_EXCL;
 
 	/*
 	 * Do the final lookup.
 	 */
-	dentry = lookup_hash(nd);
+	mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
+	dentry = lookup_hash(&nd);
 	if (IS_ERR(dentry))
 		goto fail;
 
@@ -2298,34 +2292,22 @@ struct dentry *lookup_create(struct nameidata *nd, int is_dir)
 	 * all is fine. Let's be bastards - you had / on the end, you've
 	 * been asking for (non-existent) directory. -ENOENT for you.
 	 */
-	if (unlikely(!is_dir && nd->last.name[nd->last.len])) {
+	if (unlikely(!is_dir && nd.last.name[nd.last.len])) {
 		dput(dentry);
 		dentry = ERR_PTR(-ENOENT);
+		goto fail;
 	}
+	*path = nd.path;
 	return dentry;
 eexist:
 	dput(dentry);
 	dentry = ERR_PTR(-EEXIST);
 fail:
+	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
+out:
+	path_put(&nd.path);
 	return dentry;
 }
-EXPORT_SYMBOL_GPL(lookup_create);
-
-struct dentry *kern_path_create(int dfd, const char *pathname, struct path *path, int is_dir)
-{
-	struct nameidata nd;
-	struct dentry *res;
-	int error = do_path_lookup(dfd, pathname, LOOKUP_PARENT, &nd);
-	if (error)
-		return ERR_PTR(error);
-	res = lookup_create(&nd, is_dir);
-	if (IS_ERR(res)) {
-		mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
-		path_put(&nd.path);
-	}
-	*path = nd.path;
-	return res;
-}
 EXPORT_SYMBOL(kern_path_create);
 
 struct dentry *user_path_create(int dfd, const char __user *pathname, struct path *path, int is_dir)
-- 
cgit v1.2.3


From e0a0124936171af6156b80fe8ac8799f039e767f Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 27 Jun 2011 17:00:37 -0400
Subject: switch vfs_path_lookup() to struct path

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c         | 16 +++++++++++-----
 fs/nfs/cache_lib.c |  9 ++++-----
 fs/nfs/super.c     | 16 +++++-----------
 3 files changed, 20 insertions(+), 21 deletions(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index b45a039216c7..7e6ba8c80e77 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -1575,16 +1575,22 @@ int kern_path(const char *name, unsigned int flags, struct path *path)
  * @mnt: pointer to vfs mount of the base directory
  * @name: pointer to file name
  * @flags: lookup flags
- * @nd: pointer to nameidata
+ * @path: pointer to struct path to fill
  */
 int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt,
 		    const char *name, unsigned int flags,
-		    struct nameidata *nd)
+		    struct path *path)
 {
-	nd->root.dentry = dentry;
-	nd->root.mnt = mnt;
+	struct nameidata nd;
+	int err;
+	nd.root.dentry = dentry;
+	nd.root.mnt = mnt;
+	BUG_ON(flags & LOOKUP_PARENT);
 	/* the first argument of do_path_lookup() is ignored with LOOKUP_ROOT */
-	return do_path_lookup(AT_FDCWD, name, flags | LOOKUP_ROOT, nd);
+	err = do_path_lookup(AT_FDCWD, name, flags | LOOKUP_ROOT, &nd);
+	if (!err)
+		*path = nd.path;
+	return err;
 }
 
 static struct dentry *__lookup_hash(struct qstr *name,
diff --git a/fs/nfs/cache_lib.c b/fs/nfs/cache_lib.c
index 84690319e625..c98b439332fc 100644
--- a/fs/nfs/cache_lib.c
+++ b/fs/nfs/cache_lib.c
@@ -113,19 +113,18 @@ int nfs_cache_wait_for_upcall(struct nfs_cache_defer_req *dreq)
 
 int nfs_cache_register(struct cache_detail *cd)
 {
-	struct nameidata nd;
 	struct vfsmount *mnt;
+	struct path path;
 	int ret;
 
 	mnt = rpc_get_mount();
 	if (IS_ERR(mnt))
 		return PTR_ERR(mnt);
-	ret = vfs_path_lookup(mnt->mnt_root, mnt, "/cache", 0, &nd);
+	ret = vfs_path_lookup(mnt->mnt_root, mnt, "/cache", 0, &path);
 	if (ret)
 		goto err;
-	ret = sunrpc_cache_register_pipefs(nd.path.dentry,
-			cd->name, 0600, cd);
-	path_put(&nd.path);
+	ret = sunrpc_cache_register_pipefs(path.dentry, cd->name, 0600, cd);
+	path_put(&path);
 	if (!ret)
 		return ret;
 err:
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index ce40e5c568ba..b961ceac66b4 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -2773,16 +2773,12 @@ static void nfs_referral_loop_unprotect(void)
 static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt,
 		const char *export_path)
 {
-	struct nameidata *nd = NULL;
 	struct mnt_namespace *ns_private;
 	struct super_block *s;
 	struct dentry *dentry;
+	struct path path;
 	int ret;
 
-	nd = kmalloc(sizeof(*nd), GFP_KERNEL);
-	if (nd == NULL)
-		return ERR_PTR(-ENOMEM);
-
 	ns_private = create_mnt_ns(root_mnt);
 	ret = PTR_ERR(ns_private);
 	if (IS_ERR(ns_private))
@@ -2793,7 +2789,7 @@ static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt,
 		goto out_put_mnt_ns;
 
 	ret = vfs_path_lookup(root_mnt->mnt_root, root_mnt,
-			export_path, LOOKUP_FOLLOW, nd);
+			export_path, LOOKUP_FOLLOW, &path);
 
 	nfs_referral_loop_unprotect();
 	put_mnt_ns(ns_private);
@@ -2801,12 +2797,11 @@ static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt,
 	if (ret != 0)
 		goto out_err;
 
-	s = nd->path.mnt->mnt_sb;
+	s = path.mnt->mnt_sb;
 	atomic_inc(&s->s_active);
-	dentry = dget(nd->path.dentry);
+	dentry = dget(path.dentry);
 
-	path_put(&nd->path);
-	kfree(nd);
+	path_put(&path);
 	down_write(&s->s_umount);
 	return dentry;
 out_put_mnt_ns:
@@ -2814,7 +2809,6 @@ out_put_mnt_ns:
 out_mntput:
 	mntput(root_mnt);
 out_err:
-	kfree(nd);
 	return ERR_PTR(ret);
 }
 
-- 
cgit v1.2.3


From e3c3d9c838d48c0341c40ea45ee087e3d8c8ea39 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 27 Jun 2011 17:14:56 -0400
Subject: unexport kern_path_parent()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 7e6ba8c80e77..0d188d95f244 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -3269,7 +3269,6 @@ EXPORT_SYMBOL(page_readlink);
 EXPORT_SYMBOL(__page_symlink);
 EXPORT_SYMBOL(page_symlink);
 EXPORT_SYMBOL(page_symlink_inode_operations);
-EXPORT_SYMBOL(kern_path_parent);
 EXPORT_SYMBOL(kern_path);
 EXPORT_SYMBOL(vfs_path_lookup);
 EXPORT_SYMBOL(inode_permission);
-- 
cgit v1.2.3


From a4464dbc0ca6a3ab8e9d1206bc05059dae2a559d Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 7 Jul 2011 15:03:58 -0400
Subject: Make ->d_sb assign-once and always non-NULL

New helper (non-exported, fs/internal.h-only): __d_alloc(sb, name).
Allocates dentry, sets its ->d_sb to given superblock and sets
->d_op accordingly.  Old d_alloc(NULL, name) callers are converted
to that (all of them know what superblock they want).  d_alloc()
itself is left only for parent != NULl case; uses __d_alloc(),
inserts result into the list of parent's children.

Note that now ->d_sb is assign-once and never NULL *and*
->d_parent is never NULL either.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c   | 75 +++++++++++++++++++++++++++++++----------------------------
 fs/internal.h |  5 ++++
 fs/libfs.c    |  6 ++---
 3 files changed, 47 insertions(+), 39 deletions(-)

(limited to 'fs')

diff --git a/fs/dcache.c b/fs/dcache.c
index d3902139b533..c61edd0318c3 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1275,8 +1275,8 @@ static struct shrinker dcache_shrinker = {
 };
 
 /**
- * d_alloc	-	allocate a dcache entry
- * @parent: parent of entry to allocate
+ * __d_alloc	-	allocate a dcache entry
+ * @sb: filesystem it will belong to
  * @name: qstr of the name
  *
  * Allocates a dentry. It returns %NULL if there is insufficient memory
@@ -1284,7 +1284,7 @@ static struct shrinker dcache_shrinker = {
  * copied and the copy passed in may be reused after this call.
  */
  
-struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
+struct dentry *__d_alloc(struct super_block *sb, const struct qstr *name)
 {
 	struct dentry *dentry;
 	char *dname;
@@ -1314,8 +1314,8 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
 	spin_lock_init(&dentry->d_lock);
 	seqcount_init(&dentry->d_seq);
 	dentry->d_inode = NULL;
-	dentry->d_parent = NULL;
-	dentry->d_sb = NULL;
+	dentry->d_parent = dentry;
+	dentry->d_sb = sb;
 	dentry->d_op = NULL;
 	dentry->d_fsdata = NULL;
 	INIT_HLIST_BL_NODE(&dentry->d_hash);
@@ -1323,36 +1323,47 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
 	INIT_LIST_HEAD(&dentry->d_subdirs);
 	INIT_LIST_HEAD(&dentry->d_alias);
 	INIT_LIST_HEAD(&dentry->d_u.d_child);
-
-	if (parent) {
-		spin_lock(&parent->d_lock);
-		/*
-		 * don't need child lock because it is not subject
-		 * to concurrency here
-		 */
-		__dget_dlock(parent);
-		dentry->d_parent = parent;
-		dentry->d_sb = parent->d_sb;
-		d_set_d_op(dentry, dentry->d_sb->s_d_op);
-		list_add(&dentry->d_u.d_child, &parent->d_subdirs);
-		spin_unlock(&parent->d_lock);
-	}
+	d_set_d_op(dentry, dentry->d_sb->s_d_op);
 
 	this_cpu_inc(nr_dentry);
 
 	return dentry;
 }
+
+/**
+ * d_alloc	-	allocate a dcache entry
+ * @parent: parent of entry to allocate
+ * @name: qstr of the name
+ *
+ * Allocates a dentry. It returns %NULL if there is insufficient memory
+ * available. On a success the dentry is returned. The name passed in is
+ * copied and the copy passed in may be reused after this call.
+ */
+struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
+{
+	struct dentry *dentry = __d_alloc(parent->d_sb, name);
+	if (!dentry)
+		return NULL;
+
+	spin_lock(&parent->d_lock);
+	/*
+	 * don't need child lock because it is not subject
+	 * to concurrency here
+	 */
+	__dget_dlock(parent);
+	dentry->d_parent = parent;
+	list_add(&dentry->d_u.d_child, &parent->d_subdirs);
+	spin_unlock(&parent->d_lock);
+
+	return dentry;
+}
 EXPORT_SYMBOL(d_alloc);
 
 struct dentry *d_alloc_pseudo(struct super_block *sb, const struct qstr *name)
 {
-	struct dentry *dentry = d_alloc(NULL, name);
-	if (dentry) {
-		dentry->d_sb = sb;
-		d_set_d_op(dentry, dentry->d_sb->s_d_op);
-		dentry->d_parent = dentry;
+	struct dentry *dentry = __d_alloc(sb, name);
+	if (dentry)
 		dentry->d_flags |= DCACHE_DISCONNECTED;
-	}
 	return dentry;
 }
 EXPORT_SYMBOL(d_alloc_pseudo);
@@ -1522,13 +1533,9 @@ struct dentry * d_alloc_root(struct inode * root_inode)
 	if (root_inode) {
 		static const struct qstr name = { .name = "/", .len = 1 };
 
-		res = d_alloc(NULL, &name);
-		if (res) {
-			res->d_sb = root_inode->i_sb;
-			d_set_d_op(res, res->d_sb->s_d_op);
-			res->d_parent = res;
+		res = __d_alloc(root_inode->i_sb, &name);
+		if (res)
 			d_instantiate(res, root_inode);
-		}
 	}
 	return res;
 }
@@ -1589,13 +1596,11 @@ struct dentry *d_obtain_alias(struct inode *inode)
 	if (res)
 		goto out_iput;
 
-	tmp = d_alloc(NULL, &anonstring);
+	tmp = __d_alloc(inode->i_sb, &anonstring);
 	if (!tmp) {
 		res = ERR_PTR(-ENOMEM);
 		goto out_iput;
 	}
-	tmp->d_parent = tmp; /* make sure dput doesn't croak */
-
 
 	spin_lock(&inode->i_lock);
 	res = __d_find_any_alias(inode);
@@ -1607,8 +1612,6 @@ struct dentry *d_obtain_alias(struct inode *inode)
 
 	/* attach a disconnected dentry */
 	spin_lock(&tmp->d_lock);
-	tmp->d_sb = inode->i_sb;
-	d_set_d_op(tmp, tmp->d_sb->s_d_op);
 	tmp->d_inode = inode;
 	tmp->d_flags |= DCACHE_DISCONNECTED;
 	list_add(&tmp->d_alias, &inode->i_dentry);
diff --git a/fs/internal.h b/fs/internal.h
index b29c46e4e32f..ae47c48bedde 100644
--- a/fs/internal.h
+++ b/fs/internal.h
@@ -135,3 +135,8 @@ extern void inode_wb_list_del(struct inode *inode);
 extern int get_nr_dirty_inodes(void);
 extern void evict_inodes(struct super_block *);
 extern int invalidate_inodes(struct super_block *, bool);
+
+/*
+ * dcache.c
+ */
+extern struct dentry *__d_alloc(struct super_block *, const struct qstr *);
diff --git a/fs/libfs.c b/fs/libfs.c
index 275ca4749a2e..bd50b11f92da 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -16,6 +16,8 @@
 
 #include <asm/uaccess.h>
 
+#include "internal.h"
+
 static inline int simple_positive(struct dentry *dentry)
 {
 	return dentry->d_inode && !d_unhashed(dentry);
@@ -246,13 +248,11 @@ struct dentry *mount_pseudo(struct file_system_type *fs_type, char *name,
 	root->i_ino = 1;
 	root->i_mode = S_IFDIR | S_IRUSR | S_IWUSR;
 	root->i_atime = root->i_mtime = root->i_ctime = CURRENT_TIME;
-	dentry = d_alloc(NULL, &d_name);
+	dentry = __d_alloc(s, &d_name);
 	if (!dentry) {
 		iput(root);
 		goto Enomem;
 	}
-	dentry->d_sb = s;
-	dentry->d_parent = dentry;
 	d_instantiate(dentry, root);
 	s->s_root = dentry;
 	s->s_d_op = dops;
-- 
cgit v1.2.3


From fb408e6ccc32404a05783911b6f3fed56bd17b06 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 7 Jul 2011 15:12:51 -0400
Subject: get rid of pointless checks for dentry->sb == NULL

it never is...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/cachefiles/bind.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'fs')

diff --git a/fs/cachefiles/bind.c b/fs/cachefiles/bind.c
index a2603e7c0bb5..e10c4415e8c3 100644
--- a/fs/cachefiles/bind.c
+++ b/fs/cachefiles/bind.c
@@ -129,7 +129,6 @@ static int cachefiles_daemon_add_cache(struct cachefiles_cache *cache)
 	    !root->d_inode->i_op->mkdir ||
 	    !root->d_inode->i_op->setxattr ||
 	    !root->d_inode->i_op->getxattr ||
-	    !root->d_sb ||
 	    !root->d_sb->s_op ||
 	    !root->d_sb->s_op->statfs ||
 	    !root->d_sb->s_op->sync_fs)
-- 
cgit v1.2.3


From 0ee5dc676a5f8fadede608c7281dfedb1ae714ea Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 7 Jul 2011 15:44:25 -0400
Subject: btrfs: kill magical embedded struct superblock

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/btrfs/ctree.h   |  2 +-
 fs/btrfs/disk-io.c | 15 ++++-----------
 fs/btrfs/inode.c   |  2 +-
 fs/super.c         | 32 +++++++++++++++++++++++---------
 4 files changed, 29 insertions(+), 22 deletions(-)

(limited to 'fs')

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 9864cec801ed..9552afc24ef7 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -1219,7 +1219,7 @@ struct btrfs_root {
 	 * right now this just gets used so that a root has its own devid
 	 * for stat.  It may be used for more later
 	 */
-	struct super_block anon_super;
+	dev_t anon_dev;
 };
 
 struct btrfs_ioctl_defrag_range_args {
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 1ac8db5dc0a3..b231ae13b269 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1077,12 +1077,7 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
 	init_completion(&root->kobj_unregister);
 	root->defrag_running = 0;
 	root->root_key.objectid = objectid;
-	root->anon_super.s_root = NULL;
-	root->anon_super.s_dev = 0;
-	INIT_LIST_HEAD(&root->anon_super.s_list);
-	INIT_LIST_HEAD(&root->anon_super.s_instances);
-	init_rwsem(&root->anon_super.s_umount);
-
+	root->anon_dev = 0;
 	return 0;
 }
 
@@ -1311,7 +1306,7 @@ again:
 	spin_lock_init(&root->cache_lock);
 	init_waitqueue_head(&root->cache_wait);
 
-	ret = set_anon_super(&root->anon_super, NULL);
+	ret = get_anon_bdev(&root->anon_dev);
 	if (ret)
 		goto fail;
 
@@ -2393,10 +2388,8 @@ static void free_fs_root(struct btrfs_root *root)
 {
 	iput(root->cache_inode);
 	WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
-	if (root->anon_super.s_dev) {
-		down_write(&root->anon_super.s_umount);
-		kill_anon_super(&root->anon_super);
-	}
+	if (root->anon_dev)
+		free_anon_bdev(root->anon_dev);
 	free_extent_buffer(root->node);
 	free_extent_buffer(root->commit_root);
 	kfree(root->free_ino_ctl);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index cb170ca51c64..ecf0fac712d6 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -6900,7 +6900,7 @@ static int btrfs_getattr(struct vfsmount *mnt,
 {
 	struct inode *inode = dentry->d_inode;
 	generic_fillattr(inode, stat);
-	stat->dev = BTRFS_I(inode)->root->anon_super.s_dev;
+	stat->dev = BTRFS_I(inode)->root->anon_dev;
 	stat->blksize = PAGE_CACHE_SIZE;
 	stat->blocks = (inode_get_bytes(inode) +
 			BTRFS_I(inode)->delalloc_bytes) >> 9;
diff --git a/fs/super.c b/fs/super.c
index 444da9579068..263edeb9f0e9 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -693,7 +693,7 @@ static DEFINE_IDA(unnamed_dev_ida);
 static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */
 static int unnamed_dev_start = 0; /* don't bother trying below it */
 
-int set_anon_super(struct super_block *s, void *data)
+int get_anon_bdev(dev_t *p)
 {
 	int dev;
 	int error;
@@ -720,24 +720,38 @@ int set_anon_super(struct super_block *s, void *data)
 		spin_unlock(&unnamed_dev_lock);
 		return -EMFILE;
 	}
-	s->s_dev = MKDEV(0, dev & MINORMASK);
-	s->s_bdi = &noop_backing_dev_info;
+	*p = MKDEV(0, dev & MINORMASK);
 	return 0;
 }
+EXPORT_SYMBOL(get_anon_bdev);
 
-EXPORT_SYMBOL(set_anon_super);
-
-void kill_anon_super(struct super_block *sb)
+void free_anon_bdev(dev_t dev)
 {
-	int slot = MINOR(sb->s_dev);
-
-	generic_shutdown_super(sb);
+	int slot = MINOR(dev);
 	spin_lock(&unnamed_dev_lock);
 	ida_remove(&unnamed_dev_ida, slot);
 	if (slot < unnamed_dev_start)
 		unnamed_dev_start = slot;
 	spin_unlock(&unnamed_dev_lock);
 }
+EXPORT_SYMBOL(free_anon_bdev);
+
+int set_anon_super(struct super_block *s, void *data)
+{
+	int error = get_anon_bdev(&s->s_dev);
+	if (!error)
+		s->s_bdi = &noop_backing_dev_info;
+	return error;
+}
+
+EXPORT_SYMBOL(set_anon_super);
+
+void kill_anon_super(struct super_block *sb)
+{
+	dev_t dev = sb->s_dev;
+	generic_shutdown_super(sb);
+	free_anon_bdev(dev);
+}
 
 EXPORT_SYMBOL(kill_anon_super);
 
-- 
cgit v1.2.3


From e7f59097071f2e193e900093742a4be85839f3d9 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 7 Jul 2011 15:45:59 -0400
Subject: kill useless checks for sb->s_op == NULL

never is...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/cachefiles/bind.c | 1 -
 fs/inode.c           | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/cachefiles/bind.c b/fs/cachefiles/bind.c
index e10c4415e8c3..622f4696e484 100644
--- a/fs/cachefiles/bind.c
+++ b/fs/cachefiles/bind.c
@@ -129,7 +129,6 @@ static int cachefiles_daemon_add_cache(struct cachefiles_cache *cache)
 	    !root->d_inode->i_op->mkdir ||
 	    !root->d_inode->i_op->setxattr ||
 	    !root->d_inode->i_op->getxattr ||
-	    !root->d_sb->s_op ||
 	    !root->d_sb->s_op->statfs ||
 	    !root->d_sb->s_op->sync_fs)
 		goto error_unsupported;
diff --git a/fs/inode.c b/fs/inode.c
index 43566d17d1b8..cbdcab88105f 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -1331,7 +1331,7 @@ static void iput_final(struct inode *inode)
 
 	WARN_ON(inode->i_state & I_NEW);
 
-	if (op && op->drop_inode)
+	if (op->drop_inode)
 		drop = op->drop_inode(inode);
 	else
 		drop = generic_drop_inode(inode);
-- 
cgit v1.2.3


From 5b4b299cc7b6adfb9401bf7f826a80f190b971be Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 7 Jul 2011 18:43:21 -0400
Subject: nfsd4_list_rec_dir(): don't bother with reopening rec_file

just rewind it to the beginning before vfs_readdir() and be
done with that...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfsd/nfs4recover.c | 52 +++++++++++++++++++++------------------------------
 1 file changed, 21 insertions(+), 31 deletions(-)

(limited to 'fs')

diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index ffb59ef6f82f..29d77f60585b 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -191,52 +191,42 @@ nfsd4_build_namelist(void *arg, const char *name, int namlen,
 }
 
 static int
-nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f)
+nfsd4_list_rec_dir(recdir_func *f)
 {
 	const struct cred *original_cred;
-	struct file *filp;
+	struct dentry *dir = rec_file->f_path.dentry;
 	LIST_HEAD(names);
-	struct name_list *entry;
-	struct dentry *dentry;
 	int status;
 
-	if (!rec_file)
-		return 0;
-
 	status = nfs4_save_creds(&original_cred);
 	if (status < 0)
 		return status;
 
-	filp = dentry_open(dget(dir), mntget(rec_file->f_path.mnt), O_RDONLY,
-			   current_cred());
-	status = PTR_ERR(filp);
-	if (IS_ERR(filp))
-		goto out;
-	status = vfs_readdir(filp, nfsd4_build_namelist, &names);
-	fput(filp);
+	status = vfs_llseek(rec_file, 0, SEEK_SET);
+	if (status < 0) {
+		nfs4_reset_creds(original_cred);
+		return status;
+	}
+
+	status = vfs_readdir(rec_file, nfsd4_build_namelist, &names);
 	mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_PARENT);
 	while (!list_empty(&names)) {
+		struct name_list *entry;
 		entry = list_entry(names.next, struct name_list, list);
-
-		dentry = lookup_one_len(entry->name, dir, HEXDIR_LEN-1);
-		if (IS_ERR(dentry)) {
-			status = PTR_ERR(dentry);
-			break;
+		if (!status) {
+			struct dentry *dentry;
+			dentry = lookup_one_len(entry->name, dir, HEXDIR_LEN-1);
+			if (IS_ERR(dentry)) {
+				status = PTR_ERR(dentry);
+				break;
+			}
+			status = f(dir, dentry);
+			dput(dentry);
 		}
-		status = f(dir, dentry);
-		dput(dentry);
-		if (status)
-			break;
 		list_del(&entry->list);
 		kfree(entry);
 	}
 	mutex_unlock(&dir->d_inode->i_mutex);
-out:
-	while (!list_empty(&names)) {
-		entry = list_entry(names.next, struct name_list, list);
-		list_del(&entry->list);
-		kfree(entry);
-	}
 	nfs4_reset_creds(original_cred);
 	return status;
 }
@@ -322,7 +312,7 @@ nfsd4_recdir_purge_old(void) {
 	status = mnt_want_write(rec_file->f_path.mnt);
 	if (status)
 		goto out;
-	status = nfsd4_list_rec_dir(rec_file->f_path.dentry, purge_old);
+	status = nfsd4_list_rec_dir(purge_old);
 	if (status == 0)
 		vfs_fsync(rec_file, 0);
 	mnt_drop_write(rec_file->f_path.mnt);
@@ -352,7 +342,7 @@ nfsd4_recdir_load(void) {
 	if (!rec_file)
 		return 0;
 
-	status = nfsd4_list_rec_dir(rec_file->f_path.dentry, load_recdir);
+	status = nfsd4_list_rec_dir(load_recdir);
 	if (status)
 		printk("nfsd4: failed loading clients from recovery"
 			" directory %s\n", rec_file->f_path.dentry->d_name.name);
-- 
cgit v1.2.3


From 0c1aa9a952c3608eb17bf990466f1491d1ee8b6c Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 8 Jul 2011 20:57:47 -0400
Subject: deuglify squashfs_lookup()

d_splice_alias(NULL, dentry) is equivalent to d_add(dentry, NULL), NULL
so no need for that if (inode) ... in there (or ERR_PTR(0), for that
matter)

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/squashfs/namei.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

(limited to 'fs')

diff --git a/fs/squashfs/namei.c b/fs/squashfs/namei.c
index 4bc63ac64bc0..51b36958492b 100644
--- a/fs/squashfs/namei.c
+++ b/fs/squashfs/namei.c
@@ -232,10 +232,7 @@ static struct dentry *squashfs_lookup(struct inode *dir, struct dentry *dentry,
 
 exit_lookup:
 	kfree(dire);
-	if (inode)
-		return d_splice_alias(inode, dentry);
-	d_add(dentry, inode);
-	return ERR_PTR(0);
+	return d_splice_alias(inode, dentry);
 
 data_error:
 	err = -EIO;
-- 
cgit v1.2.3


From a9049376ee05bf966bfe2b081b5071326856890a Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 8 Jul 2011 21:20:11 -0400
Subject: make d_splice_alias(ERR_PTR(err), dentry) = ERR_PTR(err)

... and simplify the living hell out of callers

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/btrfs/inode.c     |  8 +-------
 fs/dcache.c          |  3 +++
 fs/efs/namei.c       |  7 ++-----
 fs/exofs/namei.c     |  7 +------
 fs/ext2/namei.c      | 14 +++++---------
 fs/ext3/namei.c      | 14 +++++---------
 fs/ext4/namei.c      | 14 +++++---------
 fs/fat/namei_msdos.c | 29 ++++++++++-------------------
 fs/isofs/namei.c     | 11 +++--------
 fs/jffs2/dir.c       |  4 +---
 fs/jfs/namei.c       |  4 +---
 fs/logfs/dir.c       |  4 +---
 fs/nilfs2/namei.c    |  7 +------
 fs/squashfs/namei.c  |  5 -----
 fs/ufs/namei.c       |  2 --
 15 files changed, 39 insertions(+), 94 deletions(-)

(limited to 'fs')

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index ecf0fac712d6..bcb20a9a3b93 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -4079,13 +4079,7 @@ static int btrfs_dentry_delete(const struct dentry *dentry)
 static struct dentry *btrfs_lookup(struct inode *dir, struct dentry *dentry,
 				   struct nameidata *nd)
 {
-	struct inode *inode;
-
-	inode = btrfs_lookup_dentry(dir, dentry);
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-
-	return d_splice_alias(inode, dentry);
+	return d_splice_alias(btrfs_lookup_dentry(dir, dentry), dentry);
 }
 
 unsigned char btrfs_filetype_table[] = {
diff --git a/fs/dcache.c b/fs/dcache.c
index c61edd0318c3..41e2085d430b 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1652,6 +1652,9 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry)
 {
 	struct dentry *new = NULL;
 
+	if (IS_ERR(inode))
+		return ERR_CAST(inode);
+
 	if (inode && S_ISDIR(inode->i_mode)) {
 		spin_lock(&inode->i_lock);
 		new = __d_find_alias(inode, 1);
diff --git a/fs/efs/namei.c b/fs/efs/namei.c
index 1511bf9e5f80..832b10ded82f 100644
--- a/fs/efs/namei.c
+++ b/fs/efs/namei.c
@@ -60,14 +60,11 @@ static efs_ino_t efs_find_entry(struct inode *inode, const char *name, int len)
 
 struct dentry *efs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) {
 	efs_ino_t inodenum;
-	struct inode * inode = NULL;
+	struct inode *inode = NULL;
 
 	inodenum = efs_find_entry(dir, dentry->d_name.name, dentry->d_name.len);
-	if (inodenum) {
+	if (inodenum)
 		inode = efs_iget(dir->i_sb, inodenum);
-		if (IS_ERR(inode))
-			return ERR_CAST(inode);
-	}
 
 	return d_splice_alias(inode, dentry);
 }
diff --git a/fs/exofs/namei.c b/fs/exofs/namei.c
index 4d70db110cfc..b54c43775f17 100644
--- a/fs/exofs/namei.c
+++ b/fs/exofs/namei.c
@@ -55,12 +55,7 @@ static struct dentry *exofs_lookup(struct inode *dir, struct dentry *dentry,
 		return ERR_PTR(-ENAMETOOLONG);
 
 	ino = exofs_inode_by_name(dir, dentry);
-	inode = NULL;
-	if (ino) {
-		inode = exofs_iget(dir->i_sb, ino);
-		if (IS_ERR(inode))
-			return ERR_CAST(inode);
-	}
+	inode = ino ? exofs_iget(dir->i_sb, ino) : NULL;
 	return d_splice_alias(inode, dentry);
 }
 
diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c
index ed5c5d496ee9..d60b7099e2db 100644
--- a/fs/ext2/namei.c
+++ b/fs/ext2/namei.c
@@ -67,15 +67,11 @@ static struct dentry *ext2_lookup(struct inode * dir, struct dentry *dentry, str
 	inode = NULL;
 	if (ino) {
 		inode = ext2_iget(dir->i_sb, ino);
-		if (IS_ERR(inode)) {
-			if (PTR_ERR(inode) == -ESTALE) {
-				ext2_error(dir->i_sb, __func__,
-						"deleted inode referenced: %lu",
-						(unsigned long) ino);
-				return ERR_PTR(-EIO);
-			} else {
-				return ERR_CAST(inode);
-			}
+		if (inode == ERR_PTR(-ESTALE)) {
+			ext2_error(dir->i_sb, __func__,
+					"deleted inode referenced: %lu",
+					(unsigned long) ino);
+			return ERR_PTR(-EIO);
 		}
 	}
 	return d_splice_alias(inode, dentry);
diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c
index 34b6d9bfc48a..c095cf5640c7 100644
--- a/fs/ext3/namei.c
+++ b/fs/ext3/namei.c
@@ -1038,15 +1038,11 @@ static struct dentry *ext3_lookup(struct inode * dir, struct dentry *dentry, str
 			return ERR_PTR(-EIO);
 		}
 		inode = ext3_iget(dir->i_sb, ino);
-		if (IS_ERR(inode)) {
-			if (PTR_ERR(inode) == -ESTALE) {
-				ext3_error(dir->i_sb, __func__,
-						"deleted inode referenced: %lu",
-						ino);
-				return ERR_PTR(-EIO);
-			} else {
-				return ERR_CAST(inode);
-			}
+		if (inode == ERR_PTR(-ESTALE)) {
+			ext3_error(dir->i_sb, __func__,
+					"deleted inode referenced: %lu",
+					ino);
+			return ERR_PTR(-EIO);
 		}
 	}
 	return d_splice_alias(inode, dentry);
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index b754b7721f51..707d605bf769 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -1037,15 +1037,11 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, stru
 			return ERR_PTR(-EIO);
 		}
 		inode = ext4_iget(dir->i_sb, ino);
-		if (IS_ERR(inode)) {
-			if (PTR_ERR(inode) == -ESTALE) {
-				EXT4_ERROR_INODE(dir,
-						 "deleted inode referenced: %u",
-						 ino);
-				return ERR_PTR(-EIO);
-			} else {
-				return ERR_CAST(inode);
-			}
+		if (inode == ERR_PTR(-ESTALE)) {
+			EXT4_ERROR_INODE(dir,
+					 "deleted inode referenced: %u",
+					 ino);
+			return ERR_PTR(-EIO);
 		}
 	}
 	return d_splice_alias(inode, dentry);
diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c
index 3b222dafd15b..66e83b845455 100644
--- a/fs/fat/namei_msdos.c
+++ b/fs/fat/namei_msdos.c
@@ -209,29 +209,20 @@ static struct dentry *msdos_lookup(struct inode *dir, struct dentry *dentry,
 	int err;
 
 	lock_super(sb);
-
 	err = msdos_find(dir, dentry->d_name.name, dentry->d_name.len, &sinfo);
-	if (err) {
-		if (err == -ENOENT) {
-			inode = NULL;
-			goto out;
-		}
-		goto error;
-	}
-
-	inode = fat_build_inode(sb, sinfo.de, sinfo.i_pos);
-	brelse(sinfo.bh);
-	if (IS_ERR(inode)) {
-		err = PTR_ERR(inode);
-		goto error;
+	switch (err) {
+	case -ENOENT:
+		inode = NULL;
+		break;
+	case 0:
+		inode = fat_build_inode(sb, sinfo.de, sinfo.i_pos);
+		brelse(sinfo.bh);
+		break;
+	default:
+		inode = ERR_PTR(err);
 	}
-out:
 	unlock_super(sb);
 	return d_splice_alias(inode, dentry);
-
-error:
-	unlock_super(sb);
-	return ERR_PTR(err);
 }
 
 /***** Creates a directory entry (name is already formatted). */
diff --git a/fs/isofs/namei.c b/fs/isofs/namei.c
index 4fb3e8074fd4..68fa503d877e 100644
--- a/fs/isofs/namei.c
+++ b/fs/isofs/namei.c
@@ -183,14 +183,9 @@ struct dentry *isofs_lookup(struct inode *dir, struct dentry *dentry, struct nam
 				1024 + page_address(page));
 	__free_page(page);
 
-	inode = NULL;
-	if (found) {
-		inode = isofs_iget(dir->i_sb, block, offset);
-		if (IS_ERR(inode)) {
-			mutex_unlock(&sbi->s_mutex);
-			return ERR_CAST(inode);
-		}
-	}
+	inode = found ? isofs_iget(dir->i_sb, block, offset) : NULL;
+
 	mutex_unlock(&sbi->s_mutex);
+
 	return d_splice_alias(inode, dentry);
 }
diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c
index 4bca6a2e5c07..8f40ce4f1777 100644
--- a/fs/jffs2/dir.c
+++ b/fs/jffs2/dir.c
@@ -102,10 +102,8 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target,
 	mutex_unlock(&dir_f->sem);
 	if (ino) {
 		inode = jffs2_iget(dir_i->i_sb, ino);
-		if (IS_ERR(inode)) {
+		if (IS_ERR(inode))
 			printk(KERN_WARNING "iget() failed for ino #%u\n", ino);
-			return ERR_CAST(inode);
-		}
 	}
 
 	return d_splice_alias(inode, target);
diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index 1da0dc799286..247331551992 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -1481,10 +1481,8 @@ static struct dentry *jfs_lookup(struct inode *dip, struct dentry *dentry, struc
 	}
 
 	ip = jfs_iget(dip->i_sb, inum);
-	if (IS_ERR(ip)) {
+	if (IS_ERR(ip))
 		jfs_err("jfs_lookup: iget failed on inum %d", (uint) inum);
-		return ERR_CAST(ip);
-	}
 
 	return d_splice_alias(ip, dentry);
 }
diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c
index 1afae26cf236..b3ff3d894165 100644
--- a/fs/logfs/dir.c
+++ b/fs/logfs/dir.c
@@ -371,11 +371,9 @@ static struct dentry *logfs_lookup(struct inode *dir, struct dentry *dentry,
 	page_cache_release(page);
 
 	inode = logfs_iget(dir->i_sb, ino);
-	if (IS_ERR(inode)) {
+	if (IS_ERR(inode))
 		printk(KERN_ERR"LogFS: Cannot read inode #%llx for dentry (%lx, %lx)n",
 				ino, dir->i_ino, index);
-		return ERR_CAST(inode);
-	}
 	return d_splice_alias(inode, dentry);
 }
 
diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c
index 546849b3e88f..a3141990061e 100644
--- a/fs/nilfs2/namei.c
+++ b/fs/nilfs2/namei.c
@@ -72,12 +72,7 @@ nilfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd)
 		return ERR_PTR(-ENAMETOOLONG);
 
 	ino = nilfs_inode_by_name(dir, &dentry->d_name);
-	inode = NULL;
-	if (ino) {
-		inode = nilfs_iget(dir->i_sb, NILFS_I(dir)->i_root, ino);
-		if (IS_ERR(inode))
-			return ERR_CAST(inode);
-	}
+	inode = ino ? nilfs_iget(dir->i_sb, NILFS_I(dir)->i_root, ino) : NULL;
 	return d_splice_alias(inode, dentry);
 }
 
diff --git a/fs/squashfs/namei.c b/fs/squashfs/namei.c
index 51b36958492b..0682b38d7e31 100644
--- a/fs/squashfs/namei.c
+++ b/fs/squashfs/namei.c
@@ -220,11 +220,6 @@ static struct dentry *squashfs_lookup(struct inode *dir, struct dentry *dentry,
 					blk, off, ino_num);
 
 				inode = squashfs_iget(dir->i_sb, ino, ino_num);
-				if (IS_ERR(inode)) {
-					err = PTR_ERR(inode);
-					goto failed;
-				}
-
 				goto exit_lookup;
 			}
 		}
diff --git a/fs/ufs/namei.c b/fs/ufs/namei.c
index b57aab9a1184..639d49162241 100644
--- a/fs/ufs/namei.c
+++ b/fs/ufs/namei.c
@@ -59,8 +59,6 @@ static struct dentry *ufs_lookup(struct inode * dir, struct dentry *dentry, stru
 	if (ino)
 		inode = ufs_iget(dir->i_sb, ino);
 	unlock_ufs(dir->i_sb);
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
 	return d_splice_alias(inode, dentry);
 }
 
-- 
cgit v1.2.3


From fcb94f72d3e0f4f34b326c2986da8e5996daf72c Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:38 +1000
Subject: inode: convert inode_stat.nr_unused to per-cpu counters

Before we split up the inode_lru_lock, the unused inode counter
needs to be made independent of the global inode_lru_lock. Convert
it to per-cpu counters to do this.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/inode.c | 16 +++++++++++-----
 1 file changed, 11 insertions(+), 5 deletions(-)

(limited to 'fs')

diff --git a/fs/inode.c b/fs/inode.c
index cbdcab88105f..9a0361121712 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -95,6 +95,7 @@ EXPORT_SYMBOL(empty_aops);
 struct inodes_stat_t inodes_stat;
 
 static DEFINE_PER_CPU(unsigned int, nr_inodes);
+static DEFINE_PER_CPU(unsigned int, nr_unused);
 
 static struct kmem_cache *inode_cachep __read_mostly;
 
@@ -109,7 +110,11 @@ static int get_nr_inodes(void)
 
 static inline int get_nr_inodes_unused(void)
 {
-	return inodes_stat.nr_unused;
+	int i;
+	int sum = 0;
+	for_each_possible_cpu(i)
+		sum += per_cpu(nr_unused, i);
+	return sum < 0 ? 0 : sum;
 }
 
 int get_nr_dirty_inodes(void)
@@ -127,6 +132,7 @@ int proc_nr_inodes(ctl_table *table, int write,
 		   void __user *buffer, size_t *lenp, loff_t *ppos)
 {
 	inodes_stat.nr_inodes = get_nr_inodes();
+	inodes_stat.nr_unused = get_nr_inodes_unused();
 	return proc_dointvec(table, write, buffer, lenp, ppos);
 }
 #endif
@@ -340,7 +346,7 @@ static void inode_lru_list_add(struct inode *inode)
 	spin_lock(&inode_lru_lock);
 	if (list_empty(&inode->i_lru)) {
 		list_add(&inode->i_lru, &inode_lru);
-		inodes_stat.nr_unused++;
+		this_cpu_inc(nr_unused);
 	}
 	spin_unlock(&inode_lru_lock);
 }
@@ -350,7 +356,7 @@ static void inode_lru_list_del(struct inode *inode)
 	spin_lock(&inode_lru_lock);
 	if (!list_empty(&inode->i_lru)) {
 		list_del_init(&inode->i_lru);
-		inodes_stat.nr_unused--;
+		this_cpu_dec(nr_unused);
 	}
 	spin_unlock(&inode_lru_lock);
 }
@@ -656,7 +662,7 @@ static void prune_icache(int nr_to_scan)
 		    (inode->i_state & ~I_REFERENCED)) {
 			list_del_init(&inode->i_lru);
 			spin_unlock(&inode->i_lock);
-			inodes_stat.nr_unused--;
+			this_cpu_dec(nr_unused);
 			continue;
 		}
 
@@ -693,7 +699,7 @@ static void prune_icache(int nr_to_scan)
 		spin_unlock(&inode->i_lock);
 
 		list_move(&inode->i_lru, &freeable);
-		inodes_stat.nr_unused--;
+		this_cpu_dec(nr_unused);
 	}
 	if (current_is_kswapd())
 		__count_vm_events(KSWAPD_INODESTEAL, reap);
-- 
cgit v1.2.3


From 98b745c647a5a90c3c21ea43cbfad9a47b0dfad7 Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:39 +1000
Subject: inode: Make unused inode LRU per superblock

The inode unused list is currently a global LRU. This does not match
the other global filesystem cache - the dentry cache - which uses
per-superblock LRU lists. Hence we have related filesystem object
types using different LRU reclaimation schemes.

To enable a per-superblock filesystem cache shrinker, both of these
caches need to have per-sb unused object LRU lists. Hence this patch
converts the global inode LRU to per-sb LRUs.

The patch only does rudimentary per-sb propotioning in the shrinker
infrastructure, as this gets removed when the per-sb shrinker
callouts are introduced later on.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/inode.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++++++++++--------
 fs/super.c |  1 +
 2 files changed, 81 insertions(+), 11 deletions(-)

(limited to 'fs')

diff --git a/fs/inode.c b/fs/inode.c
index 9a0361121712..8c3491302e0c 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -34,7 +34,7 @@
  * inode->i_lock protects:
  *   inode->i_state, inode->i_hash, __iget()
  * inode_lru_lock protects:
- *   inode_lru, inode->i_lru
+ *   inode->i_sb->s_inode_lru, inode->i_lru
  * inode_sb_list_lock protects:
  *   sb->s_inodes, inode->i_sb_list
  * inode_wb_list_lock protects:
@@ -64,7 +64,6 @@ static unsigned int i_hash_shift __read_mostly;
 static struct hlist_head *inode_hashtable __read_mostly;
 static __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_hash_lock);
 
-static LIST_HEAD(inode_lru);
 static DEFINE_SPINLOCK(inode_lru_lock);
 
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_sb_list_lock);
@@ -345,7 +344,8 @@ static void inode_lru_list_add(struct inode *inode)
 {
 	spin_lock(&inode_lru_lock);
 	if (list_empty(&inode->i_lru)) {
-		list_add(&inode->i_lru, &inode_lru);
+		list_add(&inode->i_lru, &inode->i_sb->s_inode_lru);
+		inode->i_sb->s_nr_inodes_unused++;
 		this_cpu_inc(nr_unused);
 	}
 	spin_unlock(&inode_lru_lock);
@@ -356,6 +356,7 @@ static void inode_lru_list_del(struct inode *inode)
 	spin_lock(&inode_lru_lock);
 	if (!list_empty(&inode->i_lru)) {
 		list_del_init(&inode->i_lru);
+		inode->i_sb->s_nr_inodes_unused--;
 		this_cpu_dec(nr_unused);
 	}
 	spin_unlock(&inode_lru_lock);
@@ -628,21 +629,20 @@ static int can_unuse(struct inode *inode)
  * LRU does not have strict ordering. Hence we don't want to reclaim inodes
  * with this flag set because they are the inodes that are out of order.
  */
-static void prune_icache(int nr_to_scan)
+static void shrink_icache_sb(struct super_block *sb, int *nr_to_scan)
 {
 	LIST_HEAD(freeable);
 	int nr_scanned;
 	unsigned long reap = 0;
 
-	down_read(&iprune_sem);
 	spin_lock(&inode_lru_lock);
-	for (nr_scanned = 0; nr_scanned < nr_to_scan; nr_scanned++) {
+	for (nr_scanned = *nr_to_scan; nr_scanned >= 0; nr_scanned--) {
 		struct inode *inode;
 
-		if (list_empty(&inode_lru))
+		if (list_empty(&sb->s_inode_lru))
 			break;
 
-		inode = list_entry(inode_lru.prev, struct inode, i_lru);
+		inode = list_entry(sb->s_inode_lru.prev, struct inode, i_lru);
 
 		/*
 		 * we are inverting the inode_lru_lock/inode->i_lock here,
@@ -650,7 +650,7 @@ static void prune_icache(int nr_to_scan)
 		 * inode to the back of the list so we don't spin on it.
 		 */
 		if (!spin_trylock(&inode->i_lock)) {
-			list_move(&inode->i_lru, &inode_lru);
+			list_move(&inode->i_lru, &sb->s_inode_lru);
 			continue;
 		}
 
@@ -662,6 +662,7 @@ static void prune_icache(int nr_to_scan)
 		    (inode->i_state & ~I_REFERENCED)) {
 			list_del_init(&inode->i_lru);
 			spin_unlock(&inode->i_lock);
+			sb->s_nr_inodes_unused--;
 			this_cpu_dec(nr_unused);
 			continue;
 		}
@@ -669,7 +670,7 @@ static void prune_icache(int nr_to_scan)
 		/* recently referenced inodes get one more pass */
 		if (inode->i_state & I_REFERENCED) {
 			inode->i_state &= ~I_REFERENCED;
-			list_move(&inode->i_lru, &inode_lru);
+			list_move(&inode->i_lru, &sb->s_inode_lru);
 			spin_unlock(&inode->i_lock);
 			continue;
 		}
@@ -683,7 +684,7 @@ static void prune_icache(int nr_to_scan)
 			iput(inode);
 			spin_lock(&inode_lru_lock);
 
-			if (inode != list_entry(inode_lru.next,
+			if (inode != list_entry(sb->s_inode_lru.next,
 						struct inode, i_lru))
 				continue;	/* wrong inode or list_empty */
 			/* avoid lock inversions with trylock */
@@ -699,6 +700,7 @@ static void prune_icache(int nr_to_scan)
 		spin_unlock(&inode->i_lock);
 
 		list_move(&inode->i_lru, &freeable);
+		sb->s_nr_inodes_unused--;
 		this_cpu_dec(nr_unused);
 	}
 	if (current_is_kswapd())
@@ -706,8 +708,75 @@ static void prune_icache(int nr_to_scan)
 	else
 		__count_vm_events(PGINODESTEAL, reap);
 	spin_unlock(&inode_lru_lock);
+	*nr_to_scan = nr_scanned;
 
 	dispose_list(&freeable);
+}
+
+static void prune_icache(int count)
+{
+	struct super_block *sb, *p = NULL;
+	int w_count;
+	int unused = inodes_stat.nr_unused;
+	int prune_ratio;
+	int pruned;
+
+	if (unused == 0 || count == 0)
+		return;
+	down_read(&iprune_sem);
+	if (count >= unused)
+		prune_ratio = 1;
+	else
+		prune_ratio = unused / count;
+	spin_lock(&sb_lock);
+	list_for_each_entry(sb, &super_blocks, s_list) {
+		if (list_empty(&sb->s_instances))
+			continue;
+		if (sb->s_nr_inodes_unused == 0)
+			continue;
+		sb->s_count++;
+		/* Now, we reclaim unused dentrins with fairness.
+		 * We reclaim them same percentage from each superblock.
+		 * We calculate number of dentries to scan on this sb
+		 * as follows, but the implementation is arranged to avoid
+		 * overflows:
+		 * number of dentries to scan on this sb =
+		 * count * (number of dentries on this sb /
+		 * number of dentries in the machine)
+		 */
+		spin_unlock(&sb_lock);
+		if (prune_ratio != 1)
+			w_count = (sb->s_nr_inodes_unused / prune_ratio) + 1;
+		else
+			w_count = sb->s_nr_inodes_unused;
+		pruned = w_count;
+		/*
+		 * We need to be sure this filesystem isn't being unmounted,
+		 * otherwise we could race with generic_shutdown_super(), and
+		 * end up holding a reference to an inode while the filesystem
+		 * is unmounted.  So we try to get s_umount, and make sure
+		 * s_root isn't NULL.
+		 */
+		if (down_read_trylock(&sb->s_umount)) {
+			if ((sb->s_root != NULL) &&
+			    (!list_empty(&sb->s_dentry_lru))) {
+				shrink_icache_sb(sb, &w_count);
+				pruned -= w_count;
+			}
+			up_read(&sb->s_umount);
+		}
+		spin_lock(&sb_lock);
+		if (p)
+			__put_super(p);
+		count -= pruned;
+		p = sb;
+		/* more work left to do? */
+		if (count <= 0)
+			break;
+	}
+	if (p)
+		__put_super(p);
+	spin_unlock(&sb_lock);
 	up_read(&iprune_sem);
 }
 
diff --git a/fs/super.c b/fs/super.c
index 263edeb9f0e9..e8e6dbfefe8c 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -77,6 +77,7 @@ static struct super_block *alloc_super(struct file_system_type *type)
 		INIT_HLIST_BL_HEAD(&s->s_anon);
 		INIT_LIST_HEAD(&s->s_inodes);
 		INIT_LIST_HEAD(&s->s_dentry_lru);
+		INIT_LIST_HEAD(&s->s_inode_lru);
 		init_rwsem(&s->s_umount);
 		mutex_init(&s->s_lock);
 		lockdep_set_class(&s->s_umount, &type->s_umount_key);
-- 
cgit v1.2.3


From 09cc9fc7a7c3d872065426d7fb0f0ad6d3eb90fc Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:40 +1000
Subject: inode: move to per-sb LRU locks

With the inode LRUs moving to per-sb structures, there is no longer
a need for a global inode_lru_lock. The locking can be made more
fine-grained by moving to a per-sb LRU lock, isolating the LRU
operations of different filesytsems completely from each other.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/inode.c | 27 +++++++++++++--------------
 fs/super.c |  1 +
 2 files changed, 14 insertions(+), 14 deletions(-)

(limited to 'fs')

diff --git a/fs/inode.c b/fs/inode.c
index 8c3491302e0c..0450e25aeda0 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -33,7 +33,7 @@
  *
  * inode->i_lock protects:
  *   inode->i_state, inode->i_hash, __iget()
- * inode_lru_lock protects:
+ * inode->i_sb->s_inode_lru_lock protects:
  *   inode->i_sb->s_inode_lru, inode->i_lru
  * inode_sb_list_lock protects:
  *   sb->s_inodes, inode->i_sb_list
@@ -46,7 +46,7 @@
  *
  * inode_sb_list_lock
  *   inode->i_lock
- *     inode_lru_lock
+ *     inode->i_sb->s_inode_lru_lock
  *
  * inode_wb_list_lock
  *   inode->i_lock
@@ -64,8 +64,6 @@ static unsigned int i_hash_shift __read_mostly;
 static struct hlist_head *inode_hashtable __read_mostly;
 static __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_hash_lock);
 
-static DEFINE_SPINLOCK(inode_lru_lock);
-
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_sb_list_lock);
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_wb_list_lock);
 
@@ -342,24 +340,24 @@ EXPORT_SYMBOL(ihold);
 
 static void inode_lru_list_add(struct inode *inode)
 {
-	spin_lock(&inode_lru_lock);
+	spin_lock(&inode->i_sb->s_inode_lru_lock);
 	if (list_empty(&inode->i_lru)) {
 		list_add(&inode->i_lru, &inode->i_sb->s_inode_lru);
 		inode->i_sb->s_nr_inodes_unused++;
 		this_cpu_inc(nr_unused);
 	}
-	spin_unlock(&inode_lru_lock);
+	spin_unlock(&inode->i_sb->s_inode_lru_lock);
 }
 
 static void inode_lru_list_del(struct inode *inode)
 {
-	spin_lock(&inode_lru_lock);
+	spin_lock(&inode->i_sb->s_inode_lru_lock);
 	if (!list_empty(&inode->i_lru)) {
 		list_del_init(&inode->i_lru);
 		inode->i_sb->s_nr_inodes_unused--;
 		this_cpu_dec(nr_unused);
 	}
-	spin_unlock(&inode_lru_lock);
+	spin_unlock(&inode->i_sb->s_inode_lru_lock);
 }
 
 /**
@@ -615,7 +613,8 @@ static int can_unuse(struct inode *inode)
 
 /*
  * Scan `goal' inodes on the unused list for freeable ones. They are moved to a
- * temporary list and then are freed outside inode_lru_lock by dispose_list().
+ * temporary list and then are freed outside sb->s_inode_lru_lock by
+ * dispose_list().
  *
  * Any inodes which are pinned purely because of attached pagecache have their
  * pagecache removed.  If the inode has metadata buffers attached to
@@ -635,7 +634,7 @@ static void shrink_icache_sb(struct super_block *sb, int *nr_to_scan)
 	int nr_scanned;
 	unsigned long reap = 0;
 
-	spin_lock(&inode_lru_lock);
+	spin_lock(&sb->s_inode_lru_lock);
 	for (nr_scanned = *nr_to_scan; nr_scanned >= 0; nr_scanned--) {
 		struct inode *inode;
 
@@ -645,7 +644,7 @@ static void shrink_icache_sb(struct super_block *sb, int *nr_to_scan)
 		inode = list_entry(sb->s_inode_lru.prev, struct inode, i_lru);
 
 		/*
-		 * we are inverting the inode_lru_lock/inode->i_lock here,
+		 * we are inverting the sb->s_inode_lru_lock/inode->i_lock here,
 		 * so use a trylock. If we fail to get the lock, just move the
 		 * inode to the back of the list so we don't spin on it.
 		 */
@@ -677,12 +676,12 @@ static void shrink_icache_sb(struct super_block *sb, int *nr_to_scan)
 		if (inode_has_buffers(inode) || inode->i_data.nrpages) {
 			__iget(inode);
 			spin_unlock(&inode->i_lock);
-			spin_unlock(&inode_lru_lock);
+			spin_unlock(&sb->s_inode_lru_lock);
 			if (remove_inode_buffers(inode))
 				reap += invalidate_mapping_pages(&inode->i_data,
 								0, -1);
 			iput(inode);
-			spin_lock(&inode_lru_lock);
+			spin_lock(&sb->s_inode_lru_lock);
 
 			if (inode != list_entry(sb->s_inode_lru.next,
 						struct inode, i_lru))
@@ -707,7 +706,7 @@ static void shrink_icache_sb(struct super_block *sb, int *nr_to_scan)
 		__count_vm_events(KSWAPD_INODESTEAL, reap);
 	else
 		__count_vm_events(PGINODESTEAL, reap);
-	spin_unlock(&inode_lru_lock);
+	spin_unlock(&sb->s_inode_lru_lock);
 	*nr_to_scan = nr_scanned;
 
 	dispose_list(&freeable);
diff --git a/fs/super.c b/fs/super.c
index e8e6dbfefe8c..73ab9f9b3571 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -78,6 +78,7 @@ static struct super_block *alloc_super(struct file_system_type *type)
 		INIT_LIST_HEAD(&s->s_inodes);
 		INIT_LIST_HEAD(&s->s_dentry_lru);
 		INIT_LIST_HEAD(&s->s_inode_lru);
+		spin_lock_init(&s->s_inode_lru_lock);
 		init_rwsem(&s->s_umount);
 		mutex_init(&s->s_lock);
 		lockdep_set_class(&s->s_umount, &type->s_umount_key);
-- 
cgit v1.2.3


From 12ad3ab66103e6582ca69c0c9de18b13487eaaef Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:41 +1000
Subject: superblock: move pin_sb_for_writeback() to fs/super.c

The per-sb shrinker has the same requirement as the writeback
threads of ensuring that the superblock is usable and pinned for the
time it takes to run the work. Both need to take a passive reference
to the sb, take a read lock on the s_umount lock and then only
continue if an unmount is not in progress.

pin_sb_for_writeback() does this exactly, so move it to fs/super.c
and rename it to grab_super_passive() and exporting it via
fs/internal.h for all the VFS code to be able to use.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/fs-writeback.c | 28 +---------------------------
 fs/internal.h     |  1 +
 fs/super.c        | 33 +++++++++++++++++++++++++++++++++
 3 files changed, 35 insertions(+), 27 deletions(-)

(limited to 'fs')

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 0f015a0468de..b8c507ca42f7 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -460,32 +460,6 @@ writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
 	return ret;
 }
 
-/*
- * For background writeback the caller does not have the sb pinned
- * before calling writeback. So make sure that we do pin it, so it doesn't
- * go away while we are writing inodes from it.
- */
-static bool pin_sb_for_writeback(struct super_block *sb)
-{
-	spin_lock(&sb_lock);
-	if (list_empty(&sb->s_instances)) {
-		spin_unlock(&sb_lock);
-		return false;
-	}
-
-	sb->s_count++;
-	spin_unlock(&sb_lock);
-
-	if (down_read_trylock(&sb->s_umount)) {
-		if (sb->s_root)
-			return true;
-		up_read(&sb->s_umount);
-	}
-
-	put_super(sb);
-	return false;
-}
-
 /*
  * Write a portion of b_io inodes which belong to @sb.
  *
@@ -585,7 +559,7 @@ void writeback_inodes_wb(struct bdi_writeback *wb,
 		struct inode *inode = wb_inode(wb->b_io.prev);
 		struct super_block *sb = inode->i_sb;
 
-		if (!pin_sb_for_writeback(sb)) {
+		if (!grab_super_passive(sb)) {
 			requeue_io(inode);
 			continue;
 		}
diff --git a/fs/internal.h b/fs/internal.h
index ae47c48bedde..fe327c20af83 100644
--- a/fs/internal.h
+++ b/fs/internal.h
@@ -97,6 +97,7 @@ extern struct file *get_empty_filp(void);
  * super.c
  */
 extern int do_remount_sb(struct super_block *, int, void *, int);
+extern bool grab_super_passive(struct super_block *sb);
 extern void __put_super(struct super_block *sb);
 extern void put_super(struct super_block *sb);
 extern struct dentry *mount_fs(struct file_system_type *,
diff --git a/fs/super.c b/fs/super.c
index 73ab9f9b3571..e63c754447ce 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -242,6 +242,39 @@ static int grab_super(struct super_block *s) __releases(sb_lock)
 	return 0;
 }
 
+/*
+ *	grab_super_passive - acquire a passive reference
+ *	@s: reference we are trying to grab
+ *
+ *	Tries to acquire a passive reference. This is used in places where we
+ *	cannot take an active reference but we need to ensure that the
+ *	superblock does not go away while we are working on it. It returns
+ *	false if a reference was not gained, and returns true with the s_umount
+ *	lock held in read mode if a reference is gained. On successful return,
+ *	the caller must drop the s_umount lock and the passive reference when
+ *	done.
+ */
+bool grab_super_passive(struct super_block *sb)
+{
+	spin_lock(&sb_lock);
+	if (list_empty(&sb->s_instances)) {
+		spin_unlock(&sb_lock);
+		return false;
+	}
+
+	sb->s_count++;
+	spin_unlock(&sb_lock);
+
+	if (down_read_trylock(&sb->s_umount)) {
+		if (sb->s_root)
+			return true;
+		up_read(&sb->s_umount);
+	}
+
+	put_super(sb);
+	return false;
+}
+
 /*
  * Superblock locking.  We really ought to get rid of these two.
  */
-- 
cgit v1.2.3


From b0d40c92adafde7c2d81203ce7c1c69275f41140 Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:42 +1000
Subject: superblock: introduce per-sb cache shrinker infrastructure

With context based shrinkers, we can implement a per-superblock
shrinker that shrinks the caches attached to the superblock. We
currently have global shrinkers for the inode and dentry caches that
split up into per-superblock operations via a coarse proportioning
method that does not batch very well.  The global shrinkers also
have a dependency - dentries pin inodes - so we have to be very
careful about how we register the global shrinkers so that the
implicit call order is always correct.

With a per-sb shrinker callout, we can encode this dependency
directly into the per-sb shrinker, hence avoiding the need for
strictly ordering shrinker registrations. We also have no need for
any proportioning code for the shrinker subsystem already provides
this functionality across all shrinkers. Allowing the shrinker to
operate on a single superblock at a time means that we do less
superblock list traversals and locking and reclaim should batch more
effectively. This should result in less CPU overhead for reclaim and
potentially faster reclaim of items from each filesystem.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c | 121 ++++++------------------------------------------------------
 fs/inode.c  | 117 +++++-----------------------------------------------------
 fs/super.c  |  51 ++++++++++++++++++++++++-
 3 files changed, 71 insertions(+), 218 deletions(-)

(limited to 'fs')

diff --git a/fs/dcache.c b/fs/dcache.c
index 41e2085d430b..2762804a140d 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -743,13 +743,11 @@ static void shrink_dentry_list(struct list_head *list)
  *
  * If flags contains DCACHE_REFERENCED reference dentries will not be pruned.
  */
-static void __shrink_dcache_sb(struct super_block *sb, int *count, int flags)
+static void __shrink_dcache_sb(struct super_block *sb, int count, int flags)
 {
-	/* called from prune_dcache() and shrink_dcache_parent() */
 	struct dentry *dentry;
 	LIST_HEAD(referenced);
 	LIST_HEAD(tmp);
-	int cnt = *count;
 
 relock:
 	spin_lock(&dcache_lru_lock);
@@ -777,7 +775,7 @@ relock:
 		} else {
 			list_move_tail(&dentry->d_lru, &tmp);
 			spin_unlock(&dentry->d_lock);
-			if (!--cnt)
+			if (!--count)
 				break;
 		}
 		cond_resched_lock(&dcache_lru_lock);
@@ -787,83 +785,22 @@ relock:
 	spin_unlock(&dcache_lru_lock);
 
 	shrink_dentry_list(&tmp);
-
-	*count = cnt;
 }
 
 /**
- * prune_dcache - shrink the dcache
- * @count: number of entries to try to free
+ * prune_dcache_sb - shrink the dcache
+ * @nr_to_scan: number of entries to try to free
  *
- * Shrink the dcache. This is done when we need more memory, or simply when we
- * need to unmount something (at which point we need to unuse all dentries).
+ * Attempt to shrink the superblock dcache LRU by @nr_to_scan entries. This is
+ * done when we need more memory an called from the superblock shrinker
+ * function.
  *
- * This function may fail to free any resources if all the dentries are in use.
+ * This function may fail to free any resources if all the dentries are in
+ * use.
  */
-static void prune_dcache(int count)
+void prune_dcache_sb(struct super_block *sb, int nr_to_scan)
 {
-	struct super_block *sb, *p = NULL;
-	int w_count;
-	int unused = dentry_stat.nr_unused;
-	int prune_ratio;
-	int pruned;
-
-	if (unused == 0 || count == 0)
-		return;
-	if (count >= unused)
-		prune_ratio = 1;
-	else
-		prune_ratio = unused / count;
-	spin_lock(&sb_lock);
-	list_for_each_entry(sb, &super_blocks, s_list) {
-		if (list_empty(&sb->s_instances))
-			continue;
-		if (sb->s_nr_dentry_unused == 0)
-			continue;
-		sb->s_count++;
-		/* Now, we reclaim unused dentrins with fairness.
-		 * We reclaim them same percentage from each superblock.
-		 * We calculate number of dentries to scan on this sb
-		 * as follows, but the implementation is arranged to avoid
-		 * overflows:
-		 * number of dentries to scan on this sb =
-		 * count * (number of dentries on this sb /
-		 * number of dentries in the machine)
-		 */
-		spin_unlock(&sb_lock);
-		if (prune_ratio != 1)
-			w_count = (sb->s_nr_dentry_unused / prune_ratio) + 1;
-		else
-			w_count = sb->s_nr_dentry_unused;
-		pruned = w_count;
-		/*
-		 * We need to be sure this filesystem isn't being unmounted,
-		 * otherwise we could race with generic_shutdown_super(), and
-		 * end up holding a reference to an inode while the filesystem
-		 * is unmounted.  So we try to get s_umount, and make sure
-		 * s_root isn't NULL.
-		 */
-		if (down_read_trylock(&sb->s_umount)) {
-			if ((sb->s_root != NULL) &&
-			    (!list_empty(&sb->s_dentry_lru))) {
-				__shrink_dcache_sb(sb, &w_count,
-						DCACHE_REFERENCED);
-				pruned -= w_count;
-			}
-			up_read(&sb->s_umount);
-		}
-		spin_lock(&sb_lock);
-		if (p)
-			__put_super(p);
-		count -= pruned;
-		p = sb;
-		/* more work left to do? */
-		if (count <= 0)
-			break;
-	}
-	if (p)
-		__put_super(p);
-	spin_unlock(&sb_lock);
+	__shrink_dcache_sb(sb, nr_to_scan, DCACHE_REFERENCED);
 }
 
 /**
@@ -1238,42 +1175,10 @@ void shrink_dcache_parent(struct dentry * parent)
 	int found;
 
 	while ((found = select_parent(parent)) != 0)
-		__shrink_dcache_sb(sb, &found, 0);
+		__shrink_dcache_sb(sb, found, 0);
 }
 EXPORT_SYMBOL(shrink_dcache_parent);
 
-/*
- * Scan `sc->nr_slab_to_reclaim' dentries and return the number which remain.
- *
- * We need to avoid reentering the filesystem if the caller is performing a
- * GFP_NOFS allocation attempt.  One example deadlock is:
- *
- * ext2_new_block->getblk->GFP->shrink_dcache_memory->prune_dcache->
- * prune_one_dentry->dput->dentry_iput->iput->inode->i_sb->s_op->put_inode->
- * ext2_discard_prealloc->ext2_free_blocks->lock_super->DEADLOCK.
- *
- * In this case we return -1 to tell the caller that we baled.
- */
-static int shrink_dcache_memory(struct shrinker *shrink,
-				struct shrink_control *sc)
-{
-	int nr = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
-
-	if (nr) {
-		if (!(gfp_mask & __GFP_FS))
-			return -1;
-		prune_dcache(nr);
-	}
-
-	return (dentry_stat.nr_unused / 100) * sysctl_vfs_cache_pressure;
-}
-
-static struct shrinker dcache_shrinker = {
-	.shrink = shrink_dcache_memory,
-	.seeks = DEFAULT_SEEKS,
-};
-
 /**
  * __d_alloc	-	allocate a dcache entry
  * @sb: filesystem it will belong to
@@ -3083,8 +2988,6 @@ static void __init dcache_init(void)
 	 */
 	dentry_cache = KMEM_CACHE(dentry,
 		SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|SLAB_MEM_SPREAD);
-	
-	register_shrinker(&dcache_shrinker);
 
 	/* Hash may have been set up in dcache_init_early */
 	if (!hashdist)
diff --git a/fs/inode.c b/fs/inode.c
index 0450e25aeda0..1fdbb64a952f 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -73,7 +73,7 @@ __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_wb_list_lock);
  *
  * We don't actually need it to protect anything in the umount path,
  * but only need to cycle through it to make sure any inode that
- * prune_icache took off the LRU list has been fully torn down by the
+ * prune_icache_sb took off the LRU list has been fully torn down by the
  * time we are past evict_inodes.
  */
 static DECLARE_RWSEM(iprune_sem);
@@ -544,7 +544,7 @@ void evict_inodes(struct super_block *sb)
 	dispose_list(&dispose);
 
 	/*
-	 * Cycle through iprune_sem to make sure any inode that prune_icache
+	 * Cycle through iprune_sem to make sure any inode that prune_icache_sb
 	 * moved off the list before we took the lock has been fully torn
 	 * down.
 	 */
@@ -612,9 +612,10 @@ static int can_unuse(struct inode *inode)
 }
 
 /*
- * Scan `goal' inodes on the unused list for freeable ones. They are moved to a
- * temporary list and then are freed outside sb->s_inode_lru_lock by
- * dispose_list().
+ * Walk the superblock inode LRU for freeable inodes and attempt to free them.
+ * This is called from the superblock shrinker function with a number of inodes
+ * to trim from the LRU. Inodes to be freed are moved to a temporary list and
+ * then are freed outside inode_lock by dispose_list().
  *
  * Any inodes which are pinned purely because of attached pagecache have their
  * pagecache removed.  If the inode has metadata buffers attached to
@@ -628,14 +629,15 @@ static int can_unuse(struct inode *inode)
  * LRU does not have strict ordering. Hence we don't want to reclaim inodes
  * with this flag set because they are the inodes that are out of order.
  */
-static void shrink_icache_sb(struct super_block *sb, int *nr_to_scan)
+void prune_icache_sb(struct super_block *sb, int nr_to_scan)
 {
 	LIST_HEAD(freeable);
 	int nr_scanned;
 	unsigned long reap = 0;
 
+	down_read(&iprune_sem);
 	spin_lock(&sb->s_inode_lru_lock);
-	for (nr_scanned = *nr_to_scan; nr_scanned >= 0; nr_scanned--) {
+	for (nr_scanned = nr_to_scan; nr_scanned >= 0; nr_scanned--) {
 		struct inode *inode;
 
 		if (list_empty(&sb->s_inode_lru))
@@ -707,111 +709,11 @@ static void shrink_icache_sb(struct super_block *sb, int *nr_to_scan)
 	else
 		__count_vm_events(PGINODESTEAL, reap);
 	spin_unlock(&sb->s_inode_lru_lock);
-	*nr_to_scan = nr_scanned;
 
 	dispose_list(&freeable);
-}
-
-static void prune_icache(int count)
-{
-	struct super_block *sb, *p = NULL;
-	int w_count;
-	int unused = inodes_stat.nr_unused;
-	int prune_ratio;
-	int pruned;
-
-	if (unused == 0 || count == 0)
-		return;
-	down_read(&iprune_sem);
-	if (count >= unused)
-		prune_ratio = 1;
-	else
-		prune_ratio = unused / count;
-	spin_lock(&sb_lock);
-	list_for_each_entry(sb, &super_blocks, s_list) {
-		if (list_empty(&sb->s_instances))
-			continue;
-		if (sb->s_nr_inodes_unused == 0)
-			continue;
-		sb->s_count++;
-		/* Now, we reclaim unused dentrins with fairness.
-		 * We reclaim them same percentage from each superblock.
-		 * We calculate number of dentries to scan on this sb
-		 * as follows, but the implementation is arranged to avoid
-		 * overflows:
-		 * number of dentries to scan on this sb =
-		 * count * (number of dentries on this sb /
-		 * number of dentries in the machine)
-		 */
-		spin_unlock(&sb_lock);
-		if (prune_ratio != 1)
-			w_count = (sb->s_nr_inodes_unused / prune_ratio) + 1;
-		else
-			w_count = sb->s_nr_inodes_unused;
-		pruned = w_count;
-		/*
-		 * We need to be sure this filesystem isn't being unmounted,
-		 * otherwise we could race with generic_shutdown_super(), and
-		 * end up holding a reference to an inode while the filesystem
-		 * is unmounted.  So we try to get s_umount, and make sure
-		 * s_root isn't NULL.
-		 */
-		if (down_read_trylock(&sb->s_umount)) {
-			if ((sb->s_root != NULL) &&
-			    (!list_empty(&sb->s_dentry_lru))) {
-				shrink_icache_sb(sb, &w_count);
-				pruned -= w_count;
-			}
-			up_read(&sb->s_umount);
-		}
-		spin_lock(&sb_lock);
-		if (p)
-			__put_super(p);
-		count -= pruned;
-		p = sb;
-		/* more work left to do? */
-		if (count <= 0)
-			break;
-	}
-	if (p)
-		__put_super(p);
-	spin_unlock(&sb_lock);
 	up_read(&iprune_sem);
 }
 
-/*
- * shrink_icache_memory() will attempt to reclaim some unused inodes.  Here,
- * "unused" means that no dentries are referring to the inodes: the files are
- * not open and the dcache references to those inodes have already been
- * reclaimed.
- *
- * This function is passed the number of inodes to scan, and it returns the
- * total number of remaining possibly-reclaimable inodes.
- */
-static int shrink_icache_memory(struct shrinker *shrink,
-				struct shrink_control *sc)
-{
-	int nr = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
-
-	if (nr) {
-		/*
-		 * Nasty deadlock avoidance.  We may hold various FS locks,
-		 * and we don't want to recurse into the FS that called us
-		 * in clear_inode() and friends..
-		 */
-		if (!(gfp_mask & __GFP_FS))
-			return -1;
-		prune_icache(nr);
-	}
-	return (get_nr_inodes_unused() / 100) * sysctl_vfs_cache_pressure;
-}
-
-static struct shrinker icache_shrinker = {
-	.shrink = shrink_icache_memory,
-	.seeks = DEFAULT_SEEKS,
-};
-
 static void __wait_on_freeing_inode(struct inode *inode);
 /*
  * Called with the inode lock held.
@@ -1691,7 +1593,6 @@ void __init inode_init(void)
 					 (SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|
 					 SLAB_MEM_SPREAD),
 					 init_once);
-	register_shrinker(&icache_shrinker);
 
 	/* Hash may have been set up in inode_init_early */
 	if (!hashdist)
diff --git a/fs/super.c b/fs/super.c
index e63c754447ce..37a75410079e 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -38,6 +38,48 @@
 LIST_HEAD(super_blocks);
 DEFINE_SPINLOCK(sb_lock);
 
+/*
+ * One thing we have to be careful of with a per-sb shrinker is that we don't
+ * drop the last active reference to the superblock from within the shrinker.
+ * If that happens we could trigger unregistering the shrinker from within the
+ * shrinker path and that leads to deadlock on the shrinker_rwsem. Hence we
+ * take a passive reference to the superblock to avoid this from occurring.
+ */
+static int prune_super(struct shrinker *shrink, struct shrink_control *sc)
+{
+	struct super_block *sb;
+	int count;
+
+	sb = container_of(shrink, struct super_block, s_shrink);
+
+	/*
+	 * Deadlock avoidance.  We may hold various FS locks, and we don't want
+	 * to recurse into the FS that called us in clear_inode() and friends..
+	 */
+	if (sc->nr_to_scan && !(sc->gfp_mask & __GFP_FS))
+		return -1;
+
+	if (!grab_super_passive(sb))
+		return -1;
+
+	if (sc->nr_to_scan) {
+		/* proportion the scan between the two caches */
+		int total;
+
+		total = sb->s_nr_dentry_unused + sb->s_nr_inodes_unused + 1;
+		count = (sc->nr_to_scan * sb->s_nr_dentry_unused) / total;
+
+		/* prune dcache first as icache is pinned by it */
+		prune_dcache_sb(sb, count);
+		prune_icache_sb(sb, sc->nr_to_scan - count);
+	}
+
+	count = ((sb->s_nr_dentry_unused + sb->s_nr_inodes_unused) / 100)
+						* sysctl_vfs_cache_pressure;
+	drop_super(sb);
+	return count;
+}
+
 /**
  *	alloc_super	-	create new superblock
  *	@type:	filesystem type superblock should belong to
@@ -116,6 +158,9 @@ static struct super_block *alloc_super(struct file_system_type *type)
 		s->s_op = &default_op;
 		s->s_time_gran = 1000000000;
 		s->cleancache_poolid = -1;
+
+		s->s_shrink.seeks = DEFAULT_SEEKS;
+		s->s_shrink.shrink = prune_super;
 	}
 out:
 	return s;
@@ -183,6 +228,10 @@ void deactivate_locked_super(struct super_block *s)
 	if (atomic_dec_and_test(&s->s_active)) {
 		cleancache_flush_fs(s);
 		fs->kill_sb(s);
+
+		/* caches are now gone, we can safely kill the shrinker now */
+		unregister_shrinker(&s->s_shrink);
+
 		/*
 		 * We need to call rcu_barrier so all the delayed rcu free
 		 * inodes are flushed before we release the fs module.
@@ -311,7 +360,6 @@ void generic_shutdown_super(struct super_block *sb)
 {
 	const struct super_operations *sop = sb->s_op;
 
-
 	if (sb->s_root) {
 		shrink_dcache_for_umount(sb);
 		sync_filesystem(sb);
@@ -399,6 +447,7 @@ retry:
 	list_add(&s->s_instances, &type->fs_supers);
 	spin_unlock(&sb_lock);
 	get_filesystem(type);
+	register_shrinker(&s->s_shrink);
 	return s;
 }
 
-- 
cgit v1.2.3


From 4f8c19fdf3f97402b68f058b1c72a6c7166c9e59 Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:43 +1000
Subject: inode: remove iprune_sem

Now that we have per-sb shrinkers with a lifecycle that is a subset
of the superblock lifecycle and can reliably detect a filesystem
being unmounted, there is not longer any race condition for the
iprune_sem to protect against. Hence we can remove it.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/inode.c | 21 ---------------------
 1 file changed, 21 deletions(-)

(limited to 'fs')

diff --git a/fs/inode.c b/fs/inode.c
index 1fdbb64a952f..cf81baf1898a 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -67,17 +67,6 @@ static __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_hash_lock);
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_sb_list_lock);
 __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_wb_list_lock);
 
-/*
- * iprune_sem provides exclusion between the icache shrinking and the
- * umount path.
- *
- * We don't actually need it to protect anything in the umount path,
- * but only need to cycle through it to make sure any inode that
- * prune_icache_sb took off the LRU list has been fully torn down by the
- * time we are past evict_inodes.
- */
-static DECLARE_RWSEM(iprune_sem);
-
 /*
  * Empty aops. Can be used for the cases where the user does not
  * define any of the address_space operations.
@@ -542,14 +531,6 @@ void evict_inodes(struct super_block *sb)
 	spin_unlock(&inode_sb_list_lock);
 
 	dispose_list(&dispose);
-
-	/*
-	 * Cycle through iprune_sem to make sure any inode that prune_icache_sb
-	 * moved off the list before we took the lock has been fully torn
-	 * down.
-	 */
-	down_write(&iprune_sem);
-	up_write(&iprune_sem);
 }
 
 /**
@@ -635,7 +616,6 @@ void prune_icache_sb(struct super_block *sb, int nr_to_scan)
 	int nr_scanned;
 	unsigned long reap = 0;
 
-	down_read(&iprune_sem);
 	spin_lock(&sb->s_inode_lru_lock);
 	for (nr_scanned = nr_to_scan; nr_scanned >= 0; nr_scanned--) {
 		struct inode *inode;
@@ -711,7 +691,6 @@ void prune_icache_sb(struct super_block *sb, int nr_to_scan)
 	spin_unlock(&sb->s_inode_lru_lock);
 
 	dispose_list(&freeable);
-	up_read(&iprune_sem);
 }
 
 static void __wait_on_freeing_inode(struct inode *inode);
-- 
cgit v1.2.3


From 0e1fdafd93980eac62e778798549ce0f6073905c Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:44 +1000
Subject: superblock: add filesystem shrinker operations

Now we have a per-superblock shrinker implementation, we can add a
filesystem specific callout to it to allow filesystem internal
caches to be shrunk by the superblock shrinker.

Rather than perpetuate the multipurpose shrinker callback API (i.e.
nr_to_scan == 0 meaning "tell me how many objects freeable in the
cache), two operations will be added. The first will return the
number of objects that are freeable, the second is the actual
shrinker call.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/super.c | 45 +++++++++++++++++++++++++++++++++------------
 1 file changed, 33 insertions(+), 12 deletions(-)

(limited to 'fs')

diff --git a/fs/super.c b/fs/super.c
index 37a75410079e..5101f0544960 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -48,7 +48,8 @@ DEFINE_SPINLOCK(sb_lock);
 static int prune_super(struct shrinker *shrink, struct shrink_control *sc)
 {
 	struct super_block *sb;
-	int count;
+	int	fs_objects = 0;
+	int	total_objects;
 
 	sb = container_of(shrink, struct super_block, s_shrink);
 
@@ -62,22 +63,42 @@ static int prune_super(struct shrinker *shrink, struct shrink_control *sc)
 	if (!grab_super_passive(sb))
 		return -1;
 
-	if (sc->nr_to_scan) {
-		/* proportion the scan between the two caches */
-		int total;
+	if (sb->s_op && sb->s_op->nr_cached_objects)
+		fs_objects = sb->s_op->nr_cached_objects(sb);
+
+	total_objects = sb->s_nr_dentry_unused +
+			sb->s_nr_inodes_unused + fs_objects + 1;
 
-		total = sb->s_nr_dentry_unused + sb->s_nr_inodes_unused + 1;
-		count = (sc->nr_to_scan * sb->s_nr_dentry_unused) / total;
+	if (sc->nr_to_scan) {
+		int	dentries;
+		int	inodes;
+
+		/* proportion the scan between the caches */
+		dentries = (sc->nr_to_scan * sb->s_nr_dentry_unused) /
+							total_objects;
+		inodes = (sc->nr_to_scan * sb->s_nr_inodes_unused) /
+							total_objects;
+		if (fs_objects)
+			fs_objects = (sc->nr_to_scan * fs_objects) /
+							total_objects;
+		/*
+		 * prune the dcache first as the icache is pinned by it, then
+		 * prune the icache, followed by the filesystem specific caches
+		 */
+		prune_dcache_sb(sb, dentries);
+		prune_icache_sb(sb, inodes);
 
-		/* prune dcache first as icache is pinned by it */
-		prune_dcache_sb(sb, count);
-		prune_icache_sb(sb, sc->nr_to_scan - count);
+		if (fs_objects && sb->s_op->free_cached_objects) {
+			sb->s_op->free_cached_objects(sb, fs_objects);
+			fs_objects = sb->s_op->nr_cached_objects(sb);
+		}
+		total_objects = sb->s_nr_dentry_unused +
+				sb->s_nr_inodes_unused + fs_objects;
 	}
 
-	count = ((sb->s_nr_dentry_unused + sb->s_nr_inodes_unused) / 100)
-						* sysctl_vfs_cache_pressure;
+	total_objects = (total_objects / 100) * sysctl_vfs_cache_pressure;
 	drop_super(sb);
-	return count;
+	return total_objects;
 }
 
 /**
-- 
cgit v1.2.3


From 8ab47664d51a69ea79fe70bb07ca80664f74f76b Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:45 +1000
Subject: vfs: increase shrinker batch size

Now that the per-sb shrinker is responsible for shrinking 2 or more
caches, increase the batch size to keep econmies of scale for
shrinking each cache.  Increase the shrinker batch size to 1024
objects.

To allow for a large increase in batch size, add a conditional
reschedule to prune_icache_sb() so that we don't hold the LRU spin
lock for too long. This mirrors the behaviour of the
__shrink_dcache_sb(), and allows us to increase the batch size
without needing to worry about problems caused by long lock hold
times.

To ensure that filesystems using the per-sb shrinker callouts don't
cause problems, document that the object freeing method must
reschedule appropriately inside loops.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/super.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'fs')

diff --git a/fs/super.c b/fs/super.c
index 5101f0544960..7943f04cb3a9 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -182,6 +182,7 @@ static struct super_block *alloc_super(struct file_system_type *type)
 
 		s->s_shrink.seeks = DEFAULT_SEEKS;
 		s->s_shrink.shrink = prune_super;
+		s->s_shrink.batch = 1024;
 	}
 out:
 	return s;
-- 
cgit v1.2.3


From 8daaa83145ef1f0a146680618328dbbd0fa76939 Mon Sep 17 00:00:00 2001
From: Dave Chinner <dchinner@redhat.com>
Date: Fri, 8 Jul 2011 14:14:46 +1000
Subject: xfs: make use of new shrinker callout for the inode cache

Convert the inode reclaim shrinker to use the new per-sb shrinker
operations. This allows much bigger reclaim batches to be used, and
allows the XFS inode cache to be shrunk in proportion with the VFS
dentry and inode caches. This avoids the problem of the VFS caches
being shrunk significantly before the XFS inode cache is shrunk
resulting in imbalances in the caches during reclaim.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/xfs/linux-2.6/xfs_super.c | 26 ++++++++++------
 fs/xfs/linux-2.6/xfs_sync.c  | 71 +++++++++++++++++---------------------------
 fs/xfs/linux-2.6/xfs_sync.h  |  5 ++--
 3 files changed, 46 insertions(+), 56 deletions(-)

(limited to 'fs')

diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index a1a881e68a9a..a9c6ccff7b48 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -1025,11 +1025,6 @@ xfs_fs_put_super(
 {
 	struct xfs_mount	*mp = XFS_M(sb);
 
-	/*
-	 * Unregister the memory shrinker before we tear down the mount
-	 * structure so we don't have memory reclaim racing with us here.
-	 */
-	xfs_inode_shrinker_unregister(mp);
 	xfs_syncd_stop(mp);
 
 	/*
@@ -1416,8 +1411,6 @@ xfs_fs_fill_super(
 	if (error)
 		goto out_filestream_unmount;
 
-	xfs_inode_shrinker_register(mp);
-
 	error = xfs_mountfs(mp);
 	if (error)
 		goto out_syncd_stop;
@@ -1440,7 +1433,6 @@ xfs_fs_fill_super(
 	return 0;
 
  out_syncd_stop:
-	xfs_inode_shrinker_unregister(mp);
 	xfs_syncd_stop(mp);
  out_filestream_unmount:
 	xfs_filestream_unmount(mp);
@@ -1465,7 +1457,6 @@ xfs_fs_fill_super(
 	}
 
  fail_unmount:
-	xfs_inode_shrinker_unregister(mp);
 	xfs_syncd_stop(mp);
 
 	/*
@@ -1491,6 +1482,21 @@ xfs_fs_mount(
 	return mount_bdev(fs_type, flags, dev_name, data, xfs_fs_fill_super);
 }
 
+static int
+xfs_fs_nr_cached_objects(
+	struct super_block	*sb)
+{
+	return xfs_reclaim_inodes_count(XFS_M(sb));
+}
+
+static void
+xfs_fs_free_cached_objects(
+	struct super_block	*sb,
+	int			nr_to_scan)
+{
+	xfs_reclaim_inodes_nr(XFS_M(sb), nr_to_scan);
+}
+
 static const struct super_operations xfs_super_operations = {
 	.alloc_inode		= xfs_fs_alloc_inode,
 	.destroy_inode		= xfs_fs_destroy_inode,
@@ -1504,6 +1510,8 @@ static const struct super_operations xfs_super_operations = {
 	.statfs			= xfs_fs_statfs,
 	.remount_fs		= xfs_fs_remount,
 	.show_options		= xfs_fs_show_options,
+	.nr_cached_objects	= xfs_fs_nr_cached_objects,
+	.free_cached_objects	= xfs_fs_free_cached_objects,
 };
 
 static struct file_system_type xfs_fs_type = {
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index 8ecad5ff9f9b..9bd7e895a4e2 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -179,6 +179,8 @@ restart:
 		if (error == EFSCORRUPTED)
 			break;
 
+		cond_resched();
+
 	} while (nr_found && !done);
 
 	if (skipped) {
@@ -986,6 +988,8 @@ restart:
 
 			*nr_to_scan -= XFS_LOOKUP_BATCH;
 
+			cond_resched();
+
 		} while (nr_found && !done && *nr_to_scan > 0);
 
 		if (trylock && !done)
@@ -1003,7 +1007,7 @@ restart:
 	 * ensure that when we get more reclaimers than AGs we block rather
 	 * than spin trying to execute reclaim.
 	 */
-	if (trylock && skipped && *nr_to_scan > 0) {
+	if (skipped && (flags & SYNC_WAIT) && *nr_to_scan > 0) {
 		trylock = 0;
 		goto restart;
 	}
@@ -1021,44 +1025,38 @@ xfs_reclaim_inodes(
 }
 
 /*
- * Inode cache shrinker.
+ * Scan a certain number of inodes for reclaim.
  *
  * When called we make sure that there is a background (fast) inode reclaim in
- * progress, while we will throttle the speed of reclaim via doiing synchronous
+ * progress, while we will throttle the speed of reclaim via doing synchronous
  * reclaim of inodes. That means if we come across dirty inodes, we wait for
  * them to be cleaned, which we hope will not be very long due to the
  * background walker having already kicked the IO off on those dirty inodes.
  */
-static int
-xfs_reclaim_inode_shrink(
-	struct shrinker	*shrink,
-	struct shrink_control *sc)
+void
+xfs_reclaim_inodes_nr(
+	struct xfs_mount	*mp,
+	int			nr_to_scan)
 {
-	struct xfs_mount *mp;
-	struct xfs_perag *pag;
-	xfs_agnumber_t	ag;
-	int		reclaimable;
-	int nr_to_scan = sc->nr_to_scan;
-	gfp_t gfp_mask = sc->gfp_mask;
-
-	mp = container_of(shrink, struct xfs_mount, m_inode_shrink);
-	if (nr_to_scan) {
-		/* kick background reclaimer and push the AIL */
-		xfs_syncd_queue_reclaim(mp);
-		xfs_ail_push_all(mp->m_ail);
+	/* kick background reclaimer and push the AIL */
+	xfs_syncd_queue_reclaim(mp);
+	xfs_ail_push_all(mp->m_ail);
 
-		if (!(gfp_mask & __GFP_FS))
-			return -1;
+	xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT, &nr_to_scan);
+}
 
-		xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT,
-					&nr_to_scan);
-		/* terminate if we don't exhaust the scan */
-		if (nr_to_scan > 0)
-			return -1;
-       }
+/*
+ * Return the number of reclaimable inodes in the filesystem for
+ * the shrinker to determine how much to reclaim.
+ */
+int
+xfs_reclaim_inodes_count(
+	struct xfs_mount	*mp)
+{
+	struct xfs_perag	*pag;
+	xfs_agnumber_t		ag = 0;
+	int			reclaimable = 0;
 
-	reclaimable = 0;
-	ag = 0;
 	while ((pag = xfs_perag_get_tag(mp, ag, XFS_ICI_RECLAIM_TAG))) {
 		ag = pag->pag_agno + 1;
 		reclaimable += pag->pag_ici_reclaimable;
@@ -1067,18 +1065,3 @@ xfs_reclaim_inode_shrink(
 	return reclaimable;
 }
 
-void
-xfs_inode_shrinker_register(
-	struct xfs_mount	*mp)
-{
-	mp->m_inode_shrink.shrink = xfs_reclaim_inode_shrink;
-	mp->m_inode_shrink.seeks = DEFAULT_SEEKS;
-	register_shrinker(&mp->m_inode_shrink);
-}
-
-void
-xfs_inode_shrinker_unregister(
-	struct xfs_mount	*mp)
-{
-	unregister_shrinker(&mp->m_inode_shrink);
-}
diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h
index e3a6ad27415f..2e1568597764 100644
--- a/fs/xfs/linux-2.6/xfs_sync.h
+++ b/fs/xfs/linux-2.6/xfs_sync.h
@@ -43,6 +43,8 @@ void xfs_quiesce_attr(struct xfs_mount *mp);
 void xfs_flush_inodes(struct xfs_inode *ip);
 
 int xfs_reclaim_inodes(struct xfs_mount *mp, int mode);
+int xfs_reclaim_inodes_count(struct xfs_mount *mp);
+void xfs_reclaim_inodes_nr(struct xfs_mount *mp, int nr_to_scan);
 
 void xfs_inode_set_reclaim_tag(struct xfs_inode *ip);
 void __xfs_inode_set_reclaim_tag(struct xfs_perag *pag, struct xfs_inode *ip);
@@ -54,7 +56,4 @@ int xfs_inode_ag_iterator(struct xfs_mount *mp,
 	int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag, int flags),
 	int flags);
 
-void xfs_inode_shrinker_register(struct xfs_mount *mp);
-void xfs_inode_shrinker_unregister(struct xfs_mount *mp);
-
 #endif
-- 
cgit v1.2.3


From 8c5dc70aae29d2571c0f461d69b37e4e6e01ff4c Mon Sep 17 00:00:00 2001
From: Tobias Klauser <tklauser@distanz.ch>
Date: Fri, 1 Jul 2011 13:44:51 +0200
Subject: VFS: Fixup kerneldoc for generic_permission()

The flags parameter went away in
d749519b444db985e40b897f73ce1898b11f997e

Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namei.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'fs')

diff --git a/fs/namei.c b/fs/namei.c
index 0d188d95f244..b7fad009bbf6 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -213,7 +213,6 @@ other_perms:
  * generic_permission -  check for access rights on a Posix-like filesystem
  * @inode:	inode to check access rights for
  * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
- * @flags:	IPERM_FLAG_ flags.
  *
  * Used to check for read/write/execute permissions on a file.
  * We use "fsuid" for this, letting us set arbitrary permissions
-- 
cgit v1.2.3


From 582686915803e34adc8fdcd90bff7ca7f6a42221 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Fri, 24 Jun 2011 14:29:40 -0400
Subject: fat: remove i_alloc_sem abuse

Add a new rw_semaphore to protect bmap against truncate.  Previous
i_alloc_sem was abused for this, but it's going away in this series.

Note that we can't simply use i_mutex, given that the swapon code
calls ->bmap under it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/fat/fat.h   | 1 +
 fs/fat/file.c  | 2 ++
 fs/fat/inode.c | 6 ++++--
 3 files changed, 7 insertions(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/fat/fat.h b/fs/fat/fat.h
index 8276cc282dec..a975b4147e91 100644
--- a/fs/fat/fat.h
+++ b/fs/fat/fat.h
@@ -109,6 +109,7 @@ struct msdos_inode_info {
 	int i_attrs;		/* unused attribute bits */
 	loff_t i_pos;		/* on-disk position of directory entry or 0 */
 	struct hlist_node i_fat_hash;	/* hash by i_location */
+	struct rw_semaphore truncate_lock; /* protect bmap against truncate */
 	struct inode vfs_inode;
 };
 
diff --git a/fs/fat/file.c b/fs/fat/file.c
index 7018e1d8902d..a4a3a3c06436 100644
--- a/fs/fat/file.c
+++ b/fs/fat/file.c
@@ -429,8 +429,10 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
+		down_write(&MSDOS_I(inode)->truncate_lock);
 		truncate_setsize(inode, attr->ia_size);
 		fat_truncate_blocks(inode, attr->ia_size);
+		up_write(&MSDOS_I(inode)->truncate_lock);
 	}
 
 	setattr_copy(inode, attr);
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index cb8d8391ac0b..3decce46c38f 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -224,9 +224,9 @@ static sector_t _fat_bmap(struct address_space *mapping, sector_t block)
 	sector_t blocknr;
 
 	/* fat_get_cluster() assumes the requested blocknr isn't truncated. */
-	down_read(&mapping->host->i_alloc_sem);
+	down_read(&MSDOS_I(mapping->host)->truncate_lock);
 	blocknr = generic_block_bmap(mapping, block, fat_get_block);
-	up_read(&mapping->host->i_alloc_sem);
+	up_read(&MSDOS_I(mapping->host)->truncate_lock);
 
 	return blocknr;
 }
@@ -510,6 +510,8 @@ static struct inode *fat_alloc_inode(struct super_block *sb)
 	ei = kmem_cache_alloc(fat_inode_cachep, GFP_NOFS);
 	if (!ei)
 		return NULL;
+
+	init_rwsem(&ei->truncate_lock);
 	return &ei->vfs_inode;
 }
 
-- 
cgit v1.2.3


From 9ea7df534ed2a18157434a496a12cf073ca00c52 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Fri, 24 Jun 2011 14:29:41 -0400
Subject: ext4: Rewrite ext4_page_mkwrite() to use generic helpers

Rewrite ext4_page_mkwrite() to use __block_page_mkwrite() helper. This
removes the need of using i_alloc_sem to avoid races with truncate which
seems to be the wrong locking order according to lock ordering documented in
mm/rmap.c. Also calling ext4_da_write_begin() as used by the old code seems to
be problematic because we can decide to flush delay-allocated blocks which
will acquire s_umount semaphore - again creating unpleasant lock dependency
if not directly a deadlock.

Also add a check for frozen filesystem so that we don't busyloop in page fault
when the filesystem is frozen.

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ext4/inode.c | 106 +++++++++++++++++++++++++++++---------------------------
 1 file changed, 55 insertions(+), 51 deletions(-)

(limited to 'fs')

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index e3126c051006..bd309764557f 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5843,80 +5843,84 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	struct page *page = vmf->page;
 	loff_t size;
 	unsigned long len;
-	int ret = -EINVAL;
-	void *fsdata;
+	int ret;
 	struct file *file = vma->vm_file;
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct address_space *mapping = inode->i_mapping;
+	handle_t *handle;
+	get_block_t *get_block;
+	int retries = 0;
 
 	/*
-	 * Get i_alloc_sem to stop truncates messing with the inode. We cannot
-	 * get i_mutex because we are already holding mmap_sem.
+	 * This check is racy but catches the common case. We rely on
+	 * __block_page_mkwrite() to do a reliable check.
 	 */
-	down_read(&inode->i_alloc_sem);
-	size = i_size_read(inode);
-	if (page->mapping != mapping || size <= page_offset(page)
-	    || !PageUptodate(page)) {
-		/* page got truncated from under us? */
-		goto out_unlock;
+	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
+	/* Delalloc case is easy... */
+	if (test_opt(inode->i_sb, DELALLOC) &&
+	    !ext4_should_journal_data(inode) &&
+	    !ext4_nonda_switch(inode->i_sb)) {
+		do {
+			ret = __block_page_mkwrite(vma, vmf,
+						   ext4_da_get_block_prep);
+		} while (ret == -ENOSPC &&
+		       ext4_should_retry_alloc(inode->i_sb, &retries));
+		goto out_ret;
 	}
-	ret = 0;
 
 	lock_page(page);
-	wait_on_page_writeback(page);
-	if (PageMappedToDisk(page)) {
-		up_read(&inode->i_alloc_sem);
-		return VM_FAULT_LOCKED;
+	size = i_size_read(inode);
+	/* Page got truncated from under us? */
+	if (page->mapping != mapping || page_offset(page) > size) {
+		unlock_page(page);
+		ret = VM_FAULT_NOPAGE;
+		goto out;
 	}
 
 	if (page->index == size >> PAGE_CACHE_SHIFT)
 		len = size & ~PAGE_CACHE_MASK;
 	else
 		len = PAGE_CACHE_SIZE;
-
 	/*
-	 * return if we have all the buffers mapped. This avoid
-	 * the need to call write_begin/write_end which does a
-	 * journal_start/journal_stop which can block and take
-	 * long time
+	 * Return if we have all the buffers mapped. This avoids the need to do
+	 * journal_start/journal_stop which can block and take a long time
 	 */
 	if (page_has_buffers(page)) {
 		if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL,
 					ext4_bh_unmapped)) {
-			up_read(&inode->i_alloc_sem);
-			return VM_FAULT_LOCKED;
+			/* Wait so that we don't change page under IO */
+			wait_on_page_writeback(page);
+			ret = VM_FAULT_LOCKED;
+			goto out;
 		}
 	}
 	unlock_page(page);
-	/*
-	 * OK, we need to fill the hole... Do write_begin write_end
-	 * to do block allocation/reservation.We are not holding
-	 * inode.i__mutex here. That allow * parallel write_begin,
-	 * write_end call. lock_page prevent this from happening
-	 * on the same page though
-	 */
-	ret = mapping->a_ops->write_begin(file, mapping, page_offset(page),
-			len, AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata);
-	if (ret < 0)
-		goto out_unlock;
-	ret = mapping->a_ops->write_end(file, mapping, page_offset(page),
-			len, len, page, fsdata);
-	if (ret < 0)
-		goto out_unlock;
-	ret = 0;
-
-	/*
-	 * write_begin/end might have created a dirty page and someone
-	 * could wander in and start the IO.  Make sure that hasn't
-	 * happened.
-	 */
-	lock_page(page);
-	wait_on_page_writeback(page);
-	up_read(&inode->i_alloc_sem);
-	return VM_FAULT_LOCKED;
-out_unlock:
-	if (ret)
+	/* OK, we need to fill the hole... */
+	if (ext4_should_dioread_nolock(inode))
+		get_block = ext4_get_block_write;
+	else
+		get_block = ext4_get_block;
+retry_alloc:
+	handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode));
+	if (IS_ERR(handle)) {
 		ret = VM_FAULT_SIGBUS;
-	up_read(&inode->i_alloc_sem);
+		goto out;
+	}
+	ret = __block_page_mkwrite(vma, vmf, get_block);
+	if (!ret && ext4_should_journal_data(inode)) {
+		if (walk_page_buffers(handle, page_buffers(page), 0,
+			  PAGE_CACHE_SIZE, NULL, do_journal_get_write_access)) {
+			unlock_page(page);
+			ret = VM_FAULT_SIGBUS;
+			goto out;
+		}
+		ext4_set_inode_state(inode, EXT4_STATE_JDATA);
+	}
+	ext4_journal_stop(handle);
+	if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
+		goto retry_alloc;
+out_ret:
+	ret = block_page_mkwrite_return(ret);
+out:
 	return ret;
 }
-- 
cgit v1.2.3


From f9b5570d7fdedff32a2e78102bfb54cd1b12b289 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Fri, 24 Jun 2011 14:29:42 -0400
Subject: fs: simplify handling of zero sized reads in __blockdev_direct_IO

Reject zero sized reads as soon as we know our I/O length, and don't
borther with locks or allocations that might have to be cleaned up
otherwise.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/direct-io.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/direct-io.c b/fs/direct-io.c
index ac5f164170e3..98ce3ac0d94b 100644
--- a/fs/direct-io.c
+++ b/fs/direct-io.c
@@ -1200,6 +1200,10 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 		}
 	}
 
+	/* watch out for a 0 len io from a tricksy fs */
+	if (rw == READ && end == offset)
+		return 0;
+
 	dio = kmalloc(sizeof(*dio), GFP_KERNEL);
 	retval = -ENOMEM;
 	if (!dio)
@@ -1213,8 +1217,7 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 
 	dio->flags = flags;
 	if (dio->flags & DIO_LOCKING) {
-		/* watch out for a 0 len io from a tricksy fs */
-		if (rw == READ && end > offset) {
+		if (rw == READ) {
 			struct address_space *mapping =
 					iocb->ki_filp->f_mapping;
 
-- 
cgit v1.2.3


From bd5fe6c5eb9c548d7f07fe8f89a150bb6705e8e3 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Fri, 24 Jun 2011 14:29:43 -0400
Subject: fs: kill i_alloc_sem

i_alloc_sem is a rather special rw_semaphore.  It's the last one that may
be released by a non-owner, and it's write side is always mirrored by
real exclusion.  It's intended use it to wait for all pending direct I/O
requests to finish before starting a truncate.

Replace it with a hand-grown construct:

 - exclusion for truncates is already guaranteed by i_mutex, so it can
   simply fall way
 - the reader side is replaced by an i_dio_count member in struct inode
   that counts the number of pending direct I/O requests.  Truncate can't
   proceed as long as it's non-zero
 - when i_dio_count reaches non-zero we wake up a pending truncate using
   wake_up_bit on a new bit in i_flags
 - new references to i_dio_count can't appear while we are waiting for
   it to read zero because the direct I/O count always needs i_mutex
   (or an equivalent like XFS's i_iolock) for starting a new operation.

This scheme is much simpler, and saves the space of a spinlock_t and a
struct list_head in struct inode (typically 160 bits on a non-debug 64-bit
system).

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/attr.c           |  5 +----
 fs/direct-io.c      | 65 +++++++++++++++++++++++++++++++++++++++++------------
 fs/inode.c          |  3 +--
 fs/ntfs/file.c      |  3 +--
 fs/ntfs/inode.c     | 10 ++-------
 fs/ocfs2/aops.c     |  7 +++---
 fs/ocfs2/file.c     | 15 ++++++-------
 fs/reiserfs/xattr.c |  3 +--
 8 files changed, 67 insertions(+), 44 deletions(-)

(limited to 'fs')

diff --git a/fs/attr.c b/fs/attr.c
index caf2aa521e2b..f177ac86fa48 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -233,16 +233,13 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
 		return error;
 
 	if (ia_valid & ATTR_SIZE)
-		down_write(&dentry->d_inode->i_alloc_sem);
+		inode_dio_wait(inode);
 
 	if (inode->i_op->setattr)
 		error = inode->i_op->setattr(dentry, attr);
 	else
 		error = simple_setattr(dentry, attr);
 
-	if (ia_valid & ATTR_SIZE)
-		up_write(&dentry->d_inode->i_alloc_sem);
-
 	if (!error)
 		fsnotify_change(dentry, ia_valid);
 
diff --git a/fs/direct-io.c b/fs/direct-io.c
index 98ce3ac0d94b..354cbdbc14bd 100644
--- a/fs/direct-io.c
+++ b/fs/direct-io.c
@@ -135,6 +135,50 @@ struct dio {
 	struct page *pages[DIO_PAGES];	/* page buffer */
 };
 
+static void __inode_dio_wait(struct inode *inode)
+{
+	wait_queue_head_t *wq = bit_waitqueue(&inode->i_state, __I_DIO_WAKEUP);
+	DEFINE_WAIT_BIT(q, &inode->i_state, __I_DIO_WAKEUP);
+
+	do {
+		prepare_to_wait(wq, &q.wait, TASK_UNINTERRUPTIBLE);
+		if (atomic_read(&inode->i_dio_count))
+			schedule();
+	} while (atomic_read(&inode->i_dio_count));
+	finish_wait(wq, &q.wait);
+}
+
+/**
+ * inode_dio_wait - wait for outstanding DIO requests to finish
+ * @inode: inode to wait for
+ *
+ * Waits for all pending direct I/O requests to finish so that we can
+ * proceed with a truncate or equivalent operation.
+ *
+ * Must be called under a lock that serializes taking new references
+ * to i_dio_count, usually by inode->i_mutex.
+ */
+void inode_dio_wait(struct inode *inode)
+{
+	if (atomic_read(&inode->i_dio_count))
+		__inode_dio_wait(inode);
+}
+EXPORT_SYMBOL_GPL(inode_dio_wait);
+
+/*
+ * inode_dio_done - signal finish of a direct I/O requests
+ * @inode: inode the direct I/O happens on
+ *
+ * This is called once we've finished processing a direct I/O request,
+ * and is used to wake up callers waiting for direct I/O to be quiesced.
+ */
+void inode_dio_done(struct inode *inode)
+{
+	if (atomic_dec_and_test(&inode->i_dio_count))
+		wake_up_bit(&inode->i_state, __I_DIO_WAKEUP);
+}
+EXPORT_SYMBOL_GPL(inode_dio_done);
+
 /*
  * How many pages are in the queue?
  */
@@ -254,9 +298,7 @@ static ssize_t dio_complete(struct dio *dio, loff_t offset, ssize_t ret, bool is
 	}
 
 	if (dio->flags & DIO_LOCKING)
-		/* lockdep: non-owner release */
-		up_read_non_owner(&dio->inode->i_alloc_sem);
-
+		inode_dio_done(dio->inode);
 	return ret;
 }
 
@@ -980,9 +1022,6 @@ out:
 	return ret;
 }
 
-/*
- * Releases both i_mutex and i_alloc_sem
- */
 static ssize_t
 direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode, 
 	const struct iovec *iov, loff_t offset, unsigned long nr_segs, 
@@ -1146,15 +1185,14 @@ direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode,
  *    For writes this function is called under i_mutex and returns with
  *    i_mutex held, for reads, i_mutex is not held on entry, but it is
  *    taken and dropped again before returning.
- *    For reads and writes i_alloc_sem is taken in shared mode and released
- *    on I/O completion (which may happen asynchronously after returning to
- *    the caller).
+ *    The i_dio_count counter keeps track of the number of outstanding
+ *    direct I/O requests, and truncate waits for it to reach zero.
+ *    New references to i_dio_count must only be grabbed with i_mutex
+ *    held.
  *
  *  - if the flags value does NOT contain DIO_LOCKING we don't use any
  *    internal locking but rather rely on the filesystem to synchronize
  *    direct I/O reads/writes versus each other and truncate.
- *    For reads and writes both i_mutex and i_alloc_sem are not held on
- *    entry and are never taken.
  */
 ssize_t
 __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
@@ -1234,10 +1272,9 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 		}
 
 		/*
-		 * Will be released at I/O completion, possibly in a
-		 * different thread.
+		 * Will be decremented at I/O completion time.
 		 */
-		down_read_non_owner(&inode->i_alloc_sem);
+		atomic_inc(&inode->i_dio_count);
 	}
 
 	/*
diff --git a/fs/inode.c b/fs/inode.c
index cf81baf1898a..96c77b81167c 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -168,8 +168,7 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
 	mutex_init(&inode->i_mutex);
 	lockdep_set_class(&inode->i_mutex, &sb->s_type->i_mutex_key);
 
-	init_rwsem(&inode->i_alloc_sem);
-	lockdep_set_class(&inode->i_alloc_sem, &sb->s_type->i_alloc_sem_key);
+	atomic_set(&inode->i_dio_count, 0);
 
 	mapping->a_ops = &empty_aops;
 	mapping->host = inode;
diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c
index f4b1057abdd2..b59f5ac26bef 100644
--- a/fs/ntfs/file.c
+++ b/fs/ntfs/file.c
@@ -1832,9 +1832,8 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb,
 	 * fails again.
 	 */
 	if (unlikely(NInoTruncateFailed(ni))) {
-		down_write(&vi->i_alloc_sem);
+		inode_dio_wait(vi);
 		err = ntfs_truncate(vi);
-		up_write(&vi->i_alloc_sem);
 		if (err || NInoTruncateFailed(ni)) {
 			if (!err)
 				err = -EIO;
diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c
index c05d6dcf77a4..1371487da955 100644
--- a/fs/ntfs/inode.c
+++ b/fs/ntfs/inode.c
@@ -2357,12 +2357,7 @@ static const char *es = "  Leaving inconsistent metadata.  Unmount and run "
  *
  * Returns 0 on success or -errno on error.
  *
- * Called with ->i_mutex held.  In all but one case ->i_alloc_sem is held for
- * writing.  The only case in the kernel where ->i_alloc_sem is not held is
- * mm/filemap.c::generic_file_buffered_write() where vmtruncate() is called
- * with the current i_size as the offset.  The analogous place in NTFS is in
- * fs/ntfs/file.c::ntfs_file_buffered_write() where we call vmtruncate() again
- * without holding ->i_alloc_sem.
+ * Called with ->i_mutex held.
  */
 int ntfs_truncate(struct inode *vi)
 {
@@ -2887,8 +2882,7 @@ void ntfs_truncate_vfs(struct inode *vi) {
  * We also abort all changes of user, group, and mode as we do not implement
  * the NTFS ACLs yet.
  *
- * Called with ->i_mutex held.  For the ATTR_SIZE (i.e. ->truncate) case, also
- * called with ->i_alloc_sem held for writing.
+ * Called with ->i_mutex held.
  */
 int ntfs_setattr(struct dentry *dentry, struct iattr *attr)
 {
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index ac97bca282d2..de1d3953599d 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -551,9 +551,8 @@ bail:
 
 /*
  * ocfs2_dio_end_io is called by the dio core when a dio is finished.  We're
- * particularly interested in the aio/dio case.  Like the core uses
- * i_alloc_sem, we use the rw_lock DLM lock to protect io on one node from
- * truncation on another.
+ * particularly interested in the aio/dio case.  We use the rw_lock DLM lock
+ * to protect io on one node from truncation on another.
  */
 static void ocfs2_dio_end_io(struct kiocb *iocb,
 			     loff_t offset,
@@ -569,7 +568,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 	BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
 
 	if (ocfs2_iocb_is_sem_locked(iocb)) {
-		up_read(&inode->i_alloc_sem);
+		inode_dio_done(inode);
 		ocfs2_iocb_clear_sem_locked(iocb);
 	}
 
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 1406c37a5722..2c3a465514a2 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -2236,9 +2236,9 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
 	ocfs2_iocb_clear_sem_locked(iocb);
 
 relock:
-	/* to match setattr's i_mutex -> i_alloc_sem -> rw_lock ordering */
+	/* to match setattr's i_mutex -> rw_lock ordering */
 	if (direct_io) {
-		down_read(&inode->i_alloc_sem);
+		atomic_inc(&inode->i_dio_count);
 		have_alloc_sem = 1;
 		/* communicate with ocfs2_dio_end_io */
 		ocfs2_iocb_set_sem_locked(iocb);
@@ -2290,7 +2290,7 @@ relock:
 	 */
 	if (direct_io && !can_do_direct) {
 		ocfs2_rw_unlock(inode, rw_level);
-		up_read(&inode->i_alloc_sem);
+		inode_dio_done(inode);
 
 		have_alloc_sem = 0;
 		rw_level = -1;
@@ -2361,8 +2361,7 @@ out_dio:
 	/*
 	 * deep in g_f_a_w_n()->ocfs2_direct_IO we pass in a ocfs2_dio_end_io
 	 * function pointer which is called when o_direct io completes so that
-	 * it can unlock our rw lock.  (it's the clustered equivalent of
-	 * i_alloc_sem; protects truncate from racing with pending ios).
+	 * it can unlock our rw lock.
 	 * Unfortunately there are error cases which call end_io and others
 	 * that don't.  so we don't have to unlock the rw_lock if either an
 	 * async dio is going to do it in the future or an end_io after an
@@ -2379,7 +2378,7 @@ out:
 
 out_sems:
 	if (have_alloc_sem) {
-		up_read(&inode->i_alloc_sem);
+		inode_dio_done(inode);
 		ocfs2_iocb_clear_sem_locked(iocb);
 	}
 
@@ -2531,8 +2530,8 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
 	 * need locks to protect pending reads from racing with truncate.
 	 */
 	if (filp->f_flags & O_DIRECT) {
-		down_read(&inode->i_alloc_sem);
 		have_alloc_sem = 1;
+		atomic_inc(&inode->i_dio_count);
 		ocfs2_iocb_set_sem_locked(iocb);
 
 		ret = ocfs2_rw_lock(inode, 0);
@@ -2575,7 +2574,7 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
 
 bail:
 	if (have_alloc_sem) {
-		up_read(&inode->i_alloc_sem);
+		inode_dio_done(inode);
 		ocfs2_iocb_clear_sem_locked(iocb);
 	}
 	if (rw_level != -1)
diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
index 4ea2ab41fdee..6938d8c68d6e 100644
--- a/fs/reiserfs/xattr.c
+++ b/fs/reiserfs/xattr.c
@@ -555,11 +555,10 @@ reiserfs_xattr_set_handle(struct reiserfs_transaction_handle *th,
 
 		reiserfs_write_unlock(inode->i_sb);
 		mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_XATTR);
-		down_write(&dentry->d_inode->i_alloc_sem);
+		inode_dio_wait(dentry->d_inode);
 		reiserfs_write_lock(inode->i_sb);
 
 		err = reiserfs_setattr(dentry, &newattrs);
-		up_write(&dentry->d_inode->i_alloc_sem);
 		mutex_unlock(&dentry->d_inode->i_mutex);
 	} else
 		update_ctime(inode);
-- 
cgit v1.2.3


From 562c72aa57c36b178eacc3500a0215651eca9429 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Fri, 24 Jun 2011 14:29:45 -0400
Subject: fs: move inode_dio_wait calls into ->setattr

Let filesystems handle waiting for direct I/O requests themselves instead
of doing it beforehand.  This means filesystem-specific locks to prevent
new dio referenes from appearing can be held.  This is important to allow
generalizing i_dio_count to non-DIO_LOCKING filesystems.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/attr.c           | 3 ---
 fs/ext2/inode.c     | 2 ++
 fs/ext3/inode.c     | 3 +++
 fs/ext4/inode.c     | 2 ++
 fs/fat/file.c       | 2 ++
 fs/gfs2/bmap.c      | 2 ++
 fs/hfs/inode.c      | 2 ++
 fs/hfsplus/inode.c  | 2 ++
 fs/jfs/file.c       | 2 ++
 fs/nilfs2/inode.c   | 2 ++
 fs/ocfs2/file.c     | 2 ++
 fs/reiserfs/inode.c | 3 +++
 12 files changed, 24 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/attr.c b/fs/attr.c
index f177ac86fa48..538e27959d3f 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -232,9 +232,6 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
 	if (error)
 		return error;
 
-	if (ia_valid & ATTR_SIZE)
-		inode_dio_wait(inode);
-
 	if (inode->i_op->setattr)
 		error = inode->i_op->setattr(dentry, attr);
 	else
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index 788e09a07f7e..06e7c767ab35 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -1184,6 +1184,8 @@ static int ext2_setsize(struct inode *inode, loff_t newsize)
 	if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
 		return -EPERM;
 
+	inode_dio_wait(inode);
+
 	if (mapping_is_xip(inode->i_mapping))
 		error = xip_truncate_page(inode->i_mapping, newsize);
 	else if (test_opt(inode->i_sb, NOBH))
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 3451d23c3bae..99c28b246b89 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -3216,6 +3216,9 @@ int ext3_setattr(struct dentry *dentry, struct iattr *attr)
 		ext3_journal_stop(handle);
 	}
 
+	if (attr->ia_valid & ATTR_SIZE)
+		inode_dio_wait(inode);
+
 	if (S_ISREG(inode->i_mode) &&
 	    attr->ia_valid & ATTR_SIZE && attr->ia_size < inode->i_size) {
 		handle_t *handle;
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index bd309764557f..9ec0a2ba2502 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5351,6 +5351,8 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
 	}
 
 	if (attr->ia_valid & ATTR_SIZE) {
+		inode_dio_wait(inode);
+
 		if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
 			struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
 
diff --git a/fs/fat/file.c b/fs/fat/file.c
index a4a3a3c06436..e1587c54d3c1 100644
--- a/fs/fat/file.c
+++ b/fs/fat/file.c
@@ -397,6 +397,8 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr)
 	 * sequence.
 	 */
 	if (attr->ia_valid & ATTR_SIZE) {
+		inode_dio_wait(inode);
+
 		if (attr->ia_size > inode->i_size) {
 			error = fat_cont_expand(inode, attr->ia_size);
 			if (error || attr->ia_valid == ATTR_SIZE)
diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index e65493a8ac00..2cd0e56b8893 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -1224,6 +1224,8 @@ int gfs2_setattr_size(struct inode *inode, u64 newsize)
 	if (ret)
 		return ret;
 
+	inode_dio_wait(inode);
+
 	oldsize = inode->i_size;
 	if (newsize >= oldsize)
 		return do_grow(inode, newsize);
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index fff16c968e67..48d567cc203e 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -615,6 +615,8 @@ int hfs_inode_setattr(struct dentry *dentry, struct iattr * attr)
 
 	if ((attr->ia_valid & ATTR_SIZE) &&
 	    attr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		error = vmtruncate(inode, attr->ia_size);
 		if (error)
 			return error;
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index b248a6cfcad9..b0a0a4b621eb 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -296,6 +296,8 @@ static int hfsplus_setattr(struct dentry *dentry, struct iattr *attr)
 
 	if ((attr->ia_valid & ATTR_SIZE) &&
 	    attr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		error = vmtruncate(inode, attr->ia_size);
 		if (error)
 			return error;
diff --git a/fs/jfs/file.c b/fs/jfs/file.c
index 2f3f531f3606..9f32315acef1 100644
--- a/fs/jfs/file.c
+++ b/fs/jfs/file.c
@@ -110,6 +110,8 @@ int jfs_setattr(struct dentry *dentry, struct iattr *iattr)
 
 	if ((iattr->ia_valid & ATTR_SIZE) &&
 	    iattr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		rc = vmtruncate(inode, iattr->ia_size);
 		if (rc)
 			return rc;
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index be8664c6a825..13f113154a9f 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -778,6 +778,8 @@ int nilfs_setattr(struct dentry *dentry, struct iattr *iattr)
 
 	if ((iattr->ia_valid & ATTR_SIZE) &&
 	    iattr->ia_size != i_size_read(inode)) {
+		inode_dio_wait(inode);
+
 		err = vmtruncate(inode, iattr->ia_size);
 		if (unlikely(err))
 			goto out_err;
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 2c3a465514a2..736283ca4a4c 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1142,6 +1142,8 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
 		if (status)
 			goto bail_unlock;
 
+		inode_dio_wait(inode);
+
 		if (i_size_read(inode) > attr->ia_size) {
 			if (ocfs2_should_order_data(inode)) {
 				status = ocfs2_begin_ordered_truncate(inode,
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 4fd5bb33dbb5..dcf543d8caf1 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -3114,6 +3114,9 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr)
 			error = -EFBIG;
 			goto out;
 		}
+
+		inode_dio_wait(inode);
+
 		/* fill in hole pointers in the expanding truncate case. */
 		if (attr->ia_size > inode->i_size) {
 			error = generic_cont_expand_simple(inode, attr->ia_size);
-- 
cgit v1.2.3


From df2d6f26586f12a24f3ae5df4e236dc5c08d6eb4 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Fri, 24 Jun 2011 14:29:46 -0400
Subject: fs: always maintain i_dio_count

Maintain i_dio_count for all filesystems, not just those using DIO_LOCKING.
This these filesystems to also protect truncate against direct I/O requests
by using common code.  Right now the only non-DIO_LOCKING filesystem that
appears to do so is XFS, which uses an opencoded variant of the i_dio_count
scheme.

Behaviour doesn't change for filesystems never calling inode_dio_wait.
For ext4 behaviour changes when using the dioread_nonlock option, which
previously was missing any protection between truncate and direct I/O reads.
For ocfs2 that handcrafted i_dio_count manipulations are replaced with
the common code now enable.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/direct-io.c  | 25 +++++++++++++------------
 fs/ocfs2/aops.c |  4 +---
 fs/ocfs2/file.c | 12 +++---------
 3 files changed, 17 insertions(+), 24 deletions(-)

(limited to 'fs')

diff --git a/fs/direct-io.c b/fs/direct-io.c
index 354cbdbc14bd..0a073c7125a6 100644
--- a/fs/direct-io.c
+++ b/fs/direct-io.c
@@ -297,8 +297,7 @@ static ssize_t dio_complete(struct dio *dio, loff_t offset, ssize_t ret, bool is
 		aio_complete(dio->iocb, ret, 0);
 	}
 
-	if (dio->flags & DIO_LOCKING)
-		inode_dio_done(dio->inode);
+	inode_dio_done(dio->inode);
 	return ret;
 }
 
@@ -1185,14 +1184,16 @@ direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode,
  *    For writes this function is called under i_mutex and returns with
  *    i_mutex held, for reads, i_mutex is not held on entry, but it is
  *    taken and dropped again before returning.
- *    The i_dio_count counter keeps track of the number of outstanding
- *    direct I/O requests, and truncate waits for it to reach zero.
- *    New references to i_dio_count must only be grabbed with i_mutex
- *    held.
- *
  *  - if the flags value does NOT contain DIO_LOCKING we don't use any
  *    internal locking but rather rely on the filesystem to synchronize
  *    direct I/O reads/writes versus each other and truncate.
+ *
+ * To help with locking against truncate we incremented the i_dio_count
+ * counter before starting direct I/O, and decrement it once we are done.
+ * Truncate can wait for it to reach zero to provide exclusion.  It is
+ * expected that filesystem provide exclusion between new direct I/O
+ * and truncates.  For DIO_LOCKING filesystems this is done by i_mutex,
+ * but other filesystems need to take care of this on their own.
  */
 ssize_t
 __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
@@ -1270,13 +1271,13 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
 				goto out;
 			}
 		}
-
-		/*
-		 * Will be decremented at I/O completion time.
-		 */
-		atomic_inc(&inode->i_dio_count);
 	}
 
+	/*
+	 * Will be decremented at I/O completion time.
+	 */
+	atomic_inc(&inode->i_dio_count);
+
 	/*
 	 * For file extending writes updating i_size before data
 	 * writeouts complete can expose uninitialized blocks. So
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index de1d3953599d..524d6167fb63 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -567,10 +567,8 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 	/* this io's submitter should not have unlocked this before we could */
 	BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
 
-	if (ocfs2_iocb_is_sem_locked(iocb)) {
-		inode_dio_done(inode);
+	if (ocfs2_iocb_is_sem_locked(iocb))
 		ocfs2_iocb_clear_sem_locked(iocb);
-	}
 
 	ocfs2_iocb_clear_rw_locked(iocb);
 
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 736283ca4a4c..22d604601957 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -2240,7 +2240,6 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
 relock:
 	/* to match setattr's i_mutex -> rw_lock ordering */
 	if (direct_io) {
-		atomic_inc(&inode->i_dio_count);
 		have_alloc_sem = 1;
 		/* communicate with ocfs2_dio_end_io */
 		ocfs2_iocb_set_sem_locked(iocb);
@@ -2292,7 +2291,6 @@ relock:
 	 */
 	if (direct_io && !can_do_direct) {
 		ocfs2_rw_unlock(inode, rw_level);
-		inode_dio_done(inode);
 
 		have_alloc_sem = 0;
 		rw_level = -1;
@@ -2379,10 +2377,8 @@ out:
 		ocfs2_rw_unlock(inode, rw_level);
 
 out_sems:
-	if (have_alloc_sem) {
-		inode_dio_done(inode);
+	if (have_alloc_sem)
 		ocfs2_iocb_clear_sem_locked(iocb);
-	}
 
 	mutex_unlock(&inode->i_mutex);
 
@@ -2533,7 +2529,6 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
 	 */
 	if (filp->f_flags & O_DIRECT) {
 		have_alloc_sem = 1;
-		atomic_inc(&inode->i_dio_count);
 		ocfs2_iocb_set_sem_locked(iocb);
 
 		ret = ocfs2_rw_lock(inode, 0);
@@ -2575,10 +2570,9 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
 	}
 
 bail:
-	if (have_alloc_sem) {
-		inode_dio_done(inode);
+	if (have_alloc_sem)
 		ocfs2_iocb_clear_sem_locked(iocb);
-	}
+
 	if (rw_level != -1)
 		ocfs2_rw_unlock(inode, rw_level);
 
-- 
cgit v1.2.3


From aacfc19c626ebd3daa675652457d71019a1f583f Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Fri, 24 Jun 2011 14:29:47 -0400
Subject: fs: simplify the blockdev_direct_IO prototype

Simple filesystems always pass inode->i_sb_bdev as the block device
argument, and never need a end_io handler.  Let's simply things for
them and for my grepping activity by dropping these arguments.  The
only thing not falling into that scheme is ext4, which passes and
end_io handler without needing special flags (yet), but given how
messy the direct I/O code there is use of __blockdev_direct_IO
in one instead of two out of three cases isn't going to make a large
difference anyway.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ext2/inode.c     |  4 ++--
 fs/ext3/inode.c     |  5 ++---
 fs/ext4/inode.c     | 12 ++++++------
 fs/fat/inode.c      |  4 ++--
 fs/hfs/inode.c      |  4 ++--
 fs/hfsplus/inode.c  |  4 ++--
 fs/jfs/inode.c      |  4 ++--
 fs/nilfs2/inode.c   |  4 ++--
 fs/reiserfs/inode.c |  5 ++---
 9 files changed, 22 insertions(+), 24 deletions(-)

(limited to 'fs')

diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index 06e7c767ab35..a8a58f63f07c 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -843,8 +843,8 @@ ext2_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
 	struct inode *inode = mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev,
-				iov, offset, nr_segs, ext2_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 ext2_get_block);
 	if (ret < 0 && (rw & WRITE))
 		ext2_write_failed(mapping, offset + iov_length(iov, nr_segs));
 	return ret;
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 99c28b246b89..2978a2a17a59 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -1816,9 +1816,8 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb,
 	}
 
 retry:
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				 offset, nr_segs,
-				 ext3_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 ext3_get_block);
 	/*
 	 * In case of error extending write may have instantiated a few
 	 * blocks outside i_size. Trim these off again.
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 9ec0a2ba2502..1f35573a34e1 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3501,10 +3501,8 @@ retry:
 				 offset, nr_segs,
 				 ext4_get_block, NULL, NULL, 0);
 	else {
-		ret = blockdev_direct_IO(rw, iocb, inode,
-				 inode->i_sb->s_bdev, iov,
-				 offset, nr_segs,
-				 ext4_get_block, NULL);
+		ret = blockdev_direct_IO(rw, iocb, inode, iov,
+				 offset, nr_segs, ext4_get_block);
 
 		if (unlikely((rw & WRITE) && ret < 0)) {
 			loff_t isize = i_size_read(inode);
@@ -3748,11 +3746,13 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb,
 			EXT4_I(inode)->cur_aio_dio = iocb->private;
 		}
 
-		ret = blockdev_direct_IO(rw, iocb, inode,
+		ret = __blockdev_direct_IO(rw, iocb, inode,
 					 inode->i_sb->s_bdev, iov,
 					 offset, nr_segs,
 					 ext4_get_block_write,
-					 ext4_end_io_dio);
+					 ext4_end_io_dio,
+					 NULL,
+					 DIO_LOCKING | DIO_SKIP_HOLES);
 		if (iocb->private)
 			EXT4_I(inode)->cur_aio_dio = NULL;
 		/*
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index 3decce46c38f..5942fec22c65 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -211,8 +211,8 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb,
 	 * FAT need to use the DIO_LOCKING for avoiding the race
 	 * condition of fat_get_block() and ->truncate().
 	 */
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev,
-				 iov, offset, nr_segs, fat_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 fat_get_block);
 	if (ret < 0 && (rw & WRITE))
 		fat_write_failed(mapping, offset + iov_length(iov, nr_segs));
 
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 48d567cc203e..5e7c3f309617 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -123,8 +123,8 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_path.dentry->d_inode->i_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs, hfs_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 hfs_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index b0a0a4b621eb..5b1cb98741cc 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -119,8 +119,8 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_path.dentry->d_inode->i_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs, hfsplus_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 hfsplus_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c
index 109655904bbc..77b69b27f825 100644
--- a/fs/jfs/inode.c
+++ b/fs/jfs/inode.c
@@ -329,8 +329,8 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				offset, nr_segs, jfs_get_block, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				 jfs_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 13f113154a9f..666628b395f1 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -259,8 +259,8 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
 		return 0;
 
 	/* Needs synchronization with the cleaner */
-	size = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs, nilfs_get_block, NULL);
+	size = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				  nilfs_get_block);
 
 	/*
 	 * In case of error extending write may have instantiated a few
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index dcf543d8caf1..2922b90ceac1 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -3068,9 +3068,8 @@ static ssize_t reiserfs_direct_IO(int rw, struct kiocb *iocb,
 	struct inode *inode = file->f_mapping->host;
 	ssize_t ret;
 
-	ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov,
-				  offset, nr_segs,
-				  reiserfs_get_blocks_direct_io, NULL);
+	ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
+				  reiserfs_get_blocks_direct_io);
 
 	/*
 	 * In case of error extending write may have instantiated a few
-- 
cgit v1.2.3


From 72c5052ddc3956d847f21c2b8d55c93664a51b2c Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Fri, 24 Jun 2011 14:29:48 -0400
Subject: fs: move inode_dio_done to the end_io handler

For filesystems that delay their end_io processing we should keep our
i_dio_count until the the processing is done.  Enable this by moving
the inode_dio_done call to the end_io handler if one exist.  Note that
the actual move to the workqueue for ext4 and XFS is not done in
this patch yet, but left to the filesystem maintainers.  At least
for XFS it's not needed yet either as XFS has an internal equivalent
to i_dio_count.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/direct-io.c              | 7 ++++---
 fs/ext4/inode.c             | 5 +++++
 fs/ocfs2/aops.c             | 1 +
 fs/xfs/linux-2.6/xfs_aops.c | 3 +++
 4 files changed, 13 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/direct-io.c b/fs/direct-io.c
index 0a073c7125a6..01d2d9ef609c 100644
--- a/fs/direct-io.c
+++ b/fs/direct-io.c
@@ -293,11 +293,12 @@ static ssize_t dio_complete(struct dio *dio, loff_t offset, ssize_t ret, bool is
 	if (dio->end_io && dio->result) {
 		dio->end_io(dio->iocb, offset, transferred,
 			    dio->map_bh.b_private, ret, is_async);
-	} else if (is_async) {
-		aio_complete(dio->iocb, ret, 0);
+	} else {
+		if (is_async)
+			aio_complete(dio->iocb, ret, 0);
+		inode_dio_done(dio->inode);
 	}
 
-	inode_dio_done(dio->inode);
 	return ret;
 }
 
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 1f35573a34e1..678cde834f19 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3573,6 +3573,7 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 			    ssize_t size, void *private, int ret,
 			    bool is_async)
 {
+	struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode;
         ext4_io_end_t *io_end = iocb->private;
 	struct workqueue_struct *wq;
 	unsigned long flags;
@@ -3594,6 +3595,7 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
 out:
 		if (is_async)
 			aio_complete(iocb, ret, 0);
+		inode_dio_done(inode);
 		return;
 	}
 
@@ -3614,6 +3616,9 @@ out:
 	/* queue the work to convert unwritten extents to written */
 	queue_work(wq, &io_end->work);
 	iocb->private = NULL;
+
+	/* XXX: probably should move into the real I/O completion handler */
+	inode_dio_done(inode);
 }
 
 static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate)
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 524d6167fb63..c1efe939c774 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -577,6 +577,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
 
 	if (is_async)
 		aio_complete(iocb, ret, 0);
+	inode_dio_done(inode);
 }
 
 /*
diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index 79ce38be15a1..b3b418f519f3 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -1339,6 +1339,9 @@ xfs_end_io_direct_write(
 	} else {
 		xfs_finish_ioend_sync(ioend);
 	}
+
+	/* XXX: probably should move into the real I/O completion handler */
+	inode_dio_done(ioend->io_inode);
 }
 
 STATIC ssize_t
-- 
cgit v1.2.3


From f15146380d28b746df3c8b81b392812eb982382a Mon Sep 17 00:00:00 2001
From: Kay Sievers <kay.sievers@vrfy.org>
Date: Tue, 12 Jul 2011 20:48:39 +0200
Subject: fs: seq_file - add event counter to simplify poll() support

Moving the event counter into the dynamically allocated 'struc seq_file'
allows poll() support without the need to allocate its own tracking
structure.

All current users are switched over to use the new counter.

Requested-by: Andrew Morton akpm@linux-foundation.org
Acked-by: NeilBrown <neilb@suse.de>
Tested-by: Lucas De Marchi lucas.demarchi@profusion.mobi
Signed-off-by: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/namespace.c | 4 ++--
 fs/proc/base.c | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/namespace.c b/fs/namespace.c
index fe59bd145d21..cda50fe9250a 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -934,8 +934,8 @@ int mnt_had_events(struct proc_mounts *p)
 	int res = 0;
 
 	br_read_lock(vfsmount_lock);
-	if (p->event != ns->event) {
-		p->event = ns->event;
+	if (p->m.poll_event != ns->event) {
+		p->m.poll_event = ns->event;
 		res = 1;
 	}
 	br_read_unlock(vfsmount_lock);
diff --git a/fs/proc/base.c b/fs/proc/base.c
index be1ff932033b..3dc5e2a5cc38 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -673,7 +673,7 @@ static int mounts_open_common(struct inode *inode, struct file *file,
 	p->m.private = p;
 	p->ns = ns;
 	p->root = root;
-	p->event = ns->event;
+	p->m.poll_event = ns->event;
 
 	return 0;
 
-- 
cgit v1.2.3


From ee60498f3ef94d7ed58dcd38077e27e839e7cafc Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 16 Jul 2011 17:06:30 -0400
Subject: coda_venus_readdir(): use offsetof()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/coda/dir.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/coda/dir.c b/fs/coda/dir.c
index cd5532398c15..9d52897c1b6d 100644
--- a/fs/coda/dir.c
+++ b/fs/coda/dir.c
@@ -449,8 +449,7 @@ static int coda_venus_readdir(struct file *coda_file, void *buf,
 	struct file *host_file;
 	struct dentry *de;
 	struct venus_dirent *vdir;
-	unsigned long vdir_size =
-	    (unsigned long)(&((struct venus_dirent *)0)->d_name);
+	unsigned long vdir_size = offsetof(struct venus_dirent, d_name);
 	unsigned int type;
 	struct qstr name;
 	ino_t ino;
-- 
cgit v1.2.3


From 2def9e4ec75e6771def66a07960dd516e119ab4c Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 16 Jul 2011 17:43:09 -0400
Subject: minix_getattr(): don't bother with ->d_parent

we can find superblock easier, TYVM...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/minix/inode.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index adcdc0a4e182..e7d23e25bf1d 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -596,8 +596,7 @@ static int minix_write_inode(struct inode *inode, struct writeback_control *wbc)
 
 int minix_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat)
 {
-	struct inode *dir = dentry->d_parent->d_inode;
-	struct super_block *sb = dir->i_sb;
+	struct super_block *sb = dentry->d_sb;
 	generic_fillattr(dentry->d_inode, stat);
 	if (INODE_VERSION(dentry->d_inode) == MINIX_V1)
 		stat->blocks = (BLOCK_SIZE / 512) * V1_minix_blocks(stat->size, sb);
-- 
cgit v1.2.3


From b85fd6bdc99da917f5ae0f90f0c2d86ef9a766aa Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 17 Jul 2011 11:19:44 -0400
Subject: don't open-code parent_ino() in assorted ->readdir()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ceph/dir.c     | 2 +-
 fs/cifs/readdir.c | 2 +-
 fs/coda/dir.c     | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index b6d436f3c1ae..e8477dc51b45 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -252,7 +252,7 @@ static int ceph_readdir(struct file *filp, void *dirent, filldir_t filldir)
 		off = 1;
 	}
 	if (filp->f_pos == 1) {
-		ino_t ino = filp->f_dentry->d_parent->d_inode->i_ino;
+		ino_t ino = parent_ino(filp->f_dentry);
 		dout("readdir off 1 -> '..'\n");
 		if (filldir(dirent, "..", 2, ceph_make_fpos(0, 1),
 			    ceph_translate_ino(inode->i_sb, ino),
diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c
index 6751e745bbc6..965a3af186a1 100644
--- a/fs/cifs/readdir.c
+++ b/fs/cifs/readdir.c
@@ -796,7 +796,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
 		file->f_pos++;
 	case 1:
 		if (filldir(direntry, "..", 2, file->f_pos,
-		     file->f_path.dentry->d_parent->d_inode->i_ino, DT_DIR) < 0) {
+		     parent_ino(file->f_path.dentry), DT_DIR) < 0) {
 			cERROR(1, "Filldir for parent dir failed");
 			rc = -ENOMEM;
 			break;
diff --git a/fs/coda/dir.c b/fs/coda/dir.c
index 9d52897c1b6d..0239433f50cb 100644
--- a/fs/coda/dir.c
+++ b/fs/coda/dir.c
@@ -473,7 +473,7 @@ static int coda_venus_readdir(struct file *coda_file, void *buf,
 		coda_file->f_pos++;
 	}
 	if (coda_file->f_pos == 1) {
-		ret = filldir(buf, "..", 2, 1, de->d_parent->d_inode->i_ino, DT_DIR);
+		ret = filldir(buf, "..", 2, 1, parent_ino(de), DT_DIR);
 		if (ret < 0)
 			goto out;
 		result++;
-- 
cgit v1.2.3


From 00eacd66cd8ab5fff9df49aa3f261ad43d495434 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Sat, 16 Jul 2011 16:46:50 -0400
Subject: ext3: make ext3 mount default to barrier=1

This patch turns on barriers by default for ext3.  mount -o barrier=0
will turn them off.  Based on a patch from Chris Mason in the SuSE tree.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Eric Sandeen <sandeen@redhat.com>
Acked-by: Jan Kara <jack@suse.cz>
Acked-by: Jeff Mahoney <jeffm@suse.com>
Acked-by: Ted Ts'o <tytso@mit.edu>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ext3/super.c | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'fs')

diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index aad153ef6b78..b57ea2f91269 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -1718,6 +1718,8 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent)
 	sbi->s_resuid = le16_to_cpu(es->s_def_resuid);
 	sbi->s_resgid = le16_to_cpu(es->s_def_resgid);
 
+	/* enable barriers by default */
+	set_opt(sbi->s_mount_opt, BARRIER);
 	set_opt(sbi->s_mount_opt, RESERVATION);
 
 	if (!parse_options ((char *) data, sb, &journal_inum, &journal_devnum,
-- 
cgit v1.2.3


From b4d5b10fb2e3a4327838c07d8ebd9e350fcc133d Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@infradead.org>
Date: Sat, 16 Jul 2011 16:47:00 -0400
Subject: reiserfs: make reiserfs default to barrier=flush

Change the default reiserfs mount option to barrier=flush.  Based on a patch
from Jeff Mahoney in the SuSE tree.

Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/reiserfs/super.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'fs')

diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c
index aa91089162cb..14363b96b6af 100644
--- a/fs/reiserfs/super.c
+++ b/fs/reiserfs/super.c
@@ -1643,6 +1643,7 @@ static int reiserfs_fill_super(struct super_block *s, void *data, int silent)
 	/* Set default values for options: non-aggressive tails, RO on errors */
 	REISERFS_SB(s)->s_mount_opt |= (1 << REISERFS_SMALLTAIL);
 	REISERFS_SB(s)->s_mount_opt |= (1 << REISERFS_ERROR_RO);
+	REISERFS_SB(s)->s_mount_opt |= (1 << REISERFS_BARRIER_FLUSH);
 	/* no preallocation minimum, be smart in
 	   reiserfs_file_write instead */
 	REISERFS_SB(s)->s_alloc_options.preallocmin = 0;
-- 
cgit v1.2.3


From 982d816581eeeacfe5b2b7c6d47d13a157616eff Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Mon, 18 Jul 2011 13:21:35 -0400
Subject: fs: add SEEK_HOLE and SEEK_DATA flags

This just gets us ready to support the SEEK_HOLE and SEEK_DATA flags.  Turns out
using fiemap in things like cp cause more problems than it solves, so lets try
and give userspace an interface that doesn't suck.  We need to match solaris
here, and the definitions are

*o* If /whence/ is SEEK_HOLE, the offset of the start of the
next hole greater than or equal to the supplied offset
is returned. The definition of a hole is provided near
the end of the DESCRIPTION.

*o* If /whence/ is SEEK_DATA, the file pointer is set to the
start of the next non-hole file region greater than or
equal to the supplied offset.

So in the generic case the entire file is data and there is a virtual hole at
the end.  That means we will just return i_size for SEEK_HOLE and will return
the same offset for SEEK_DATA.  This is how Solaris does it so we have to do it
the same way.

Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/read_write.c | 44 +++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 41 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/read_write.c b/fs/read_write.c
index 5520f8ad5504..5907b49e4d7e 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -64,6 +64,23 @@ generic_file_llseek_unlocked(struct file *file, loff_t offset, int origin)
 			return file->f_pos;
 		offset += file->f_pos;
 		break;
+	case SEEK_DATA:
+		/*
+		 * In the generic case the entire file is data, so as long as
+		 * offset isn't at the end of the file then the offset is data.
+		 */
+		if (offset >= inode->i_size)
+			return -ENXIO;
+		break;
+	case SEEK_HOLE:
+		/*
+		 * There is a virtual hole at the end of the file, so as long as
+		 * offset isn't i_size or larger, return i_size.
+		 */
+		if (offset >= inode->i_size)
+			return -ENXIO;
+		offset = inode->i_size;
+		break;
 	}
 
 	if (offset < 0 && !unsigned_offsets(file))
@@ -128,12 +145,13 @@ EXPORT_SYMBOL(no_llseek);
 
 loff_t default_llseek(struct file *file, loff_t offset, int origin)
 {
+	struct inode *inode = file->f_path.dentry->d_inode;
 	loff_t retval;
 
-	mutex_lock(&file->f_dentry->d_inode->i_mutex);
+	mutex_lock(&inode->i_mutex);
 	switch (origin) {
 		case SEEK_END:
-			offset += i_size_read(file->f_path.dentry->d_inode);
+			offset += i_size_read(inode);
 			break;
 		case SEEK_CUR:
 			if (offset == 0) {
@@ -141,6 +159,26 @@ loff_t default_llseek(struct file *file, loff_t offset, int origin)
 				goto out;
 			}
 			offset += file->f_pos;
+			break;
+		case SEEK_DATA:
+			/*
+			 * In the generic case the entire file is data, so as
+			 * long as offset isn't at the end of the file then the
+			 * offset is data.
+			 */
+			if (offset >= inode->i_size)
+				return -ENXIO;
+			break;
+		case SEEK_HOLE:
+			/*
+			 * There is a virtual hole at the end of the file, so
+			 * as long as offset isn't i_size or larger, return
+			 * i_size.
+			 */
+			if (offset >= inode->i_size)
+				return -ENXIO;
+			offset = inode->i_size;
+			break;
 	}
 	retval = -EINVAL;
 	if (offset >= 0 || unsigned_offsets(file)) {
@@ -151,7 +189,7 @@ loff_t default_llseek(struct file *file, loff_t offset, int origin)
 		retval = offset;
 	}
 out:
-	mutex_unlock(&file->f_dentry->d_inode->i_mutex);
+	mutex_unlock(&inode->i_mutex);
 	return retval;
 }
 EXPORT_SYMBOL(default_llseek);
-- 
cgit v1.2.3


From b26751575a9aa55fd6dbf3febde3ff06dfadc44f Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Mon, 18 Jul 2011 13:21:36 -0400
Subject: Btrfs: implement our own ->llseek

In order to handle SEEK_HOLE/SEEK_DATA we need to implement our own llseek.
Basically for the normal SEEK_*'s we will just defer to the generic helper, and
for SEEK_HOLE/SEEK_DATA we will use our fiemap helper to figure out the nearest
hole or data.  Currently this helper doesn't check for delalloc bytes for
prealloc space, so for now treat prealloc as data until that is fixed.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/btrfs/ctree.h |   3 ++
 fs/btrfs/file.c  | 148 ++++++++++++++++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 150 insertions(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 9552afc24ef7..f1ff62bff1b3 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2510,6 +2510,9 @@ int btrfs_csum_truncate(struct btrfs_trans_handle *trans,
 int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
 			     struct list_head *list, int search_commit);
 /* inode.c */
+struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *page,
+					   size_t pg_offset, u64 start, u64 len,
+					   int create);
 
 /* RHEL and EL kernels have a patch that renames PG_checked to FsMisc */
 #if defined(ClearPageFsMisc) && !defined(ClearPageChecked)
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index fa4ef18b66b1..bd4d061c6e4d 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1664,8 +1664,154 @@ out:
 	return ret;
 }
 
+static int find_desired_extent(struct inode *inode, loff_t *offset, int origin)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	struct extent_map *em;
+	struct extent_state *cached_state = NULL;
+	u64 lockstart = *offset;
+	u64 lockend = i_size_read(inode);
+	u64 start = *offset;
+	u64 orig_start = *offset;
+	u64 len = i_size_read(inode);
+	u64 last_end = 0;
+	int ret = 0;
+
+	lockend = max_t(u64, root->sectorsize, lockend);
+	if (lockend <= lockstart)
+		lockend = lockstart + root->sectorsize;
+
+	len = lockend - lockstart + 1;
+
+	len = max_t(u64, len, root->sectorsize);
+	if (inode->i_size == 0)
+		return -ENXIO;
+
+	lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend, 0,
+			 &cached_state, GFP_NOFS);
+
+	/*
+	 * Delalloc is such a pain.  If we have a hole and we have pending
+	 * delalloc for a portion of the hole we will get back a hole that
+	 * exists for the entire range since it hasn't been actually written
+	 * yet.  So to take care of this case we need to look for an extent just
+	 * before the position we want in case there is outstanding delalloc
+	 * going on here.
+	 */
+	if (origin == SEEK_HOLE && start != 0) {
+		if (start <= root->sectorsize)
+			em = btrfs_get_extent_fiemap(inode, NULL, 0, 0,
+						     root->sectorsize, 0);
+		else
+			em = btrfs_get_extent_fiemap(inode, NULL, 0,
+						     start - root->sectorsize,
+						     root->sectorsize, 0);
+		if (IS_ERR(em)) {
+			ret = -ENXIO;
+			goto out;
+		}
+		last_end = em->start + em->len;
+		if (em->block_start == EXTENT_MAP_DELALLOC)
+			last_end = min_t(u64, last_end, inode->i_size);
+		free_extent_map(em);
+	}
+
+	while (1) {
+		em = btrfs_get_extent_fiemap(inode, NULL, 0, start, len, 0);
+		if (IS_ERR(em)) {
+			ret = -ENXIO;
+			break;
+		}
+
+		if (em->block_start == EXTENT_MAP_HOLE) {
+			if (test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
+				if (last_end <= orig_start) {
+					free_extent_map(em);
+					ret = -ENXIO;
+					break;
+				}
+			}
+
+			if (origin == SEEK_HOLE) {
+				*offset = start;
+				free_extent_map(em);
+				break;
+			}
+		} else {
+			if (origin == SEEK_DATA) {
+				if (em->block_start == EXTENT_MAP_DELALLOC) {
+					if (start >= inode->i_size) {
+						free_extent_map(em);
+						ret = -ENXIO;
+						break;
+					}
+				}
+
+				*offset = start;
+				free_extent_map(em);
+				break;
+			}
+		}
+
+		start = em->start + em->len;
+		last_end = em->start + em->len;
+
+		if (em->block_start == EXTENT_MAP_DELALLOC)
+			last_end = min_t(u64, last_end, inode->i_size);
+
+		if (test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
+			free_extent_map(em);
+			ret = -ENXIO;
+			break;
+		}
+		free_extent_map(em);
+		cond_resched();
+	}
+	if (!ret)
+		*offset = min(*offset, inode->i_size);
+out:
+	unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend,
+			     &cached_state, GFP_NOFS);
+	return ret;
+}
+
+static loff_t btrfs_file_llseek(struct file *file, loff_t offset, int origin)
+{
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	mutex_lock(&inode->i_mutex);
+	switch (origin) {
+	case SEEK_END:
+	case SEEK_CUR:
+		offset = generic_file_llseek_unlocked(file, offset, origin);
+		goto out;
+	case SEEK_DATA:
+	case SEEK_HOLE:
+		ret = find_desired_extent(inode, &offset, origin);
+		if (ret) {
+			mutex_unlock(&inode->i_mutex);
+			return ret;
+		}
+	}
+
+	if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET))
+		return -EINVAL;
+	if (offset > inode->i_sb->s_maxbytes)
+		return -EINVAL;
+
+	/* Special lock needed here? */
+	if (offset != file->f_pos) {
+		file->f_pos = offset;
+		file->f_version = 0;
+	}
+out:
+	mutex_unlock(&inode->i_mutex);
+	return offset;
+}
+
 const struct file_operations btrfs_file_operations = {
-	.llseek		= generic_file_llseek,
+	.llseek		= btrfs_file_llseek,
 	.read		= do_sync_read,
 	.write		= do_sync_write,
 	.aio_read       = generic_file_aio_read,
-- 
cgit v1.2.3


From c334b1138bd44bea578eab7971c59bd9212a1093 Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Mon, 18 Jul 2011 13:21:37 -0400
Subject: Ext4: handle SEEK_HOLE/SEEK_DATA generically

Since Ext4 has its own lseek we need to make sure it handles
SEEK_HOLE/SEEK_DATA.  For now just do the same thing that is done in the generic
case, somebody else can come along and make it do fancy things later.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ext4/file.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

(limited to 'fs')

diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 2c0972322009..ce766f974b1d 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -236,6 +236,27 @@ loff_t ext4_llseek(struct file *file, loff_t offset, int origin)
 		}
 		offset += file->f_pos;
 		break;
+	case SEEK_DATA:
+		/*
+		 * In the generic case the entire file is data, so as long as
+		 * offset isn't at the end of the file then the offset is data.
+		 */
+		if (offset >= inode->i_size) {
+			mutex_unlock(&inode->i_mutex);
+			return -ENXIO;
+		}
+		break;
+	case SEEK_HOLE:
+		/*
+		 * There is a virtual hole at the end of the file, so as long as
+		 * offset isn't i_size or larger, return i_size.
+		 */
+		if (offset >= inode->i_size) {
+			mutex_unlock(&inode->i_mutex);
+			return -ENXIO;
+		}
+		offset = inode->i_size;
+		break;
 	}
 
 	if (offset < 0 || offset > maxbytes) {
-- 
cgit v1.2.3


From 06222e491e663dac939f04b125c9dc52126a75c4 Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Mon, 18 Jul 2011 13:21:38 -0400
Subject: fs: handle SEEK_HOLE/SEEK_DATA properly in all fs's that define their
 own llseek

This converts everybody to handle SEEK_HOLE/SEEK_DATA properly.  In some cases
we just return -EINVAL, in others we do the normal generic thing, and in others
we're simply making sure that the properly due-dilligence is done.  For example
in NFS/CIFS we need to make sure the file size is update properly for the
SEEK_HOLE and SEEK_DATA case, but since it calls the generic llseek stuff itself
that is all we have to do.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/block_dev.c   | 11 ++++++++---
 fs/ceph/dir.c    |  8 +++++++-
 fs/ceph/file.c   | 20 ++++++++++++++++++--
 fs/cifs/cifsfs.c |  7 +++++--
 fs/fuse/file.c   | 21 +++++++++++++++++++--
 fs/hpfs/dir.c    |  4 ++++
 fs/nfs/file.c    |  7 +++++--
 7 files changed, 66 insertions(+), 12 deletions(-)

(limited to 'fs')

diff --git a/fs/block_dev.c b/fs/block_dev.c
index 610e8e0b04b8..966617a422d9 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -355,20 +355,25 @@ static loff_t block_llseek(struct file *file, loff_t offset, int origin)
 	mutex_lock(&bd_inode->i_mutex);
 	size = i_size_read(bd_inode);
 
+	retval = -EINVAL;
 	switch (origin) {
-		case 2:
+		case SEEK_END:
 			offset += size;
 			break;
-		case 1:
+		case SEEK_CUR:
 			offset += file->f_pos;
+		case SEEK_SET:
+			break;
+		default:
+			goto out;
 	}
-	retval = -EINVAL;
 	if (offset >= 0 && offset <= size) {
 		if (offset != file->f_pos) {
 			file->f_pos = offset;
 		}
 		retval = offset;
 	}
+out:
 	mutex_unlock(&bd_inode->i_mutex);
 	return retval;
 }
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index e8477dc51b45..0972b457a03f 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -446,14 +446,19 @@ static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int origin)
 	loff_t retval;
 
 	mutex_lock(&inode->i_mutex);
+	retval = -EINVAL;
 	switch (origin) {
 	case SEEK_END:
 		offset += inode->i_size + 2;   /* FIXME */
 		break;
 	case SEEK_CUR:
 		offset += file->f_pos;
+	case SEEK_SET:
+		break;
+	default:
+		goto out;
 	}
-	retval = -EINVAL;
+
 	if (offset >= 0 && offset <= inode->i_sb->s_maxbytes) {
 		if (offset != file->f_pos) {
 			file->f_pos = offset;
@@ -477,6 +482,7 @@ static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int origin)
 		if (offset > old_offset)
 			fi->dir_release_count--;
 	}
+out:
 	mutex_unlock(&inode->i_mutex);
 	return retval;
 }
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 0a292459c895..0d0eae05598f 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -768,13 +768,16 @@ static loff_t ceph_llseek(struct file *file, loff_t offset, int origin)
 
 	mutex_lock(&inode->i_mutex);
 	__ceph_do_pending_vmtruncate(inode);
-	switch (origin) {
-	case SEEK_END:
+	if (origin != SEEK_CUR || origin != SEEK_SET) {
 		ret = ceph_do_getattr(inode, CEPH_STAT_CAP_SIZE);
 		if (ret < 0) {
 			offset = ret;
 			goto out;
 		}
+	}
+
+	switch (origin) {
+	case SEEK_END:
 		offset += inode->i_size;
 		break;
 	case SEEK_CUR:
@@ -790,6 +793,19 @@ static loff_t ceph_llseek(struct file *file, loff_t offset, int origin)
 		}
 		offset += file->f_pos;
 		break;
+	case SEEK_DATA:
+		if (offset >= inode->i_size) {
+			ret = -ENXIO;
+			goto out;
+		}
+		break;
+	case SEEK_HOLE:
+		if (offset >= inode->i_size) {
+			ret = -ENXIO;
+			goto out;
+		}
+		offset = inode->i_size;
+		break;
 	}
 
 	if (offset < 0 || offset > inode->i_sb->s_maxbytes) {
diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c
index cbbb55e781d3..865517470967 100644
--- a/fs/cifs/cifsfs.c
+++ b/fs/cifs/cifsfs.c
@@ -704,8 +704,11 @@ static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
 
 static loff_t cifs_llseek(struct file *file, loff_t offset, int origin)
 {
-	/* origin == SEEK_END => we must revalidate the cached file length */
-	if (origin == SEEK_END) {
+	/*
+	 * origin == SEEK_END || SEEK_DATA || SEEK_HOLE => we must revalidate
+	 * the cached file length
+	 */
+	if (origin != SEEK_SET || origin != SEEK_CUR) {
 		int rc;
 		struct inode *inode = file->f_path.dentry->d_inode;
 
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index 82a66466a24c..73b89df20851 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -1600,15 +1600,32 @@ static loff_t fuse_file_llseek(struct file *file, loff_t offset, int origin)
 	struct inode *inode = file->f_path.dentry->d_inode;
 
 	mutex_lock(&inode->i_mutex);
-	switch (origin) {
-	case SEEK_END:
+	if (origin != SEEK_CUR || origin != SEEK_SET) {
 		retval = fuse_update_attributes(inode, NULL, file, NULL);
 		if (retval)
 			goto exit;
+	}
+
+	switch (origin) {
+	case SEEK_END:
 		offset += i_size_read(inode);
 		break;
 	case SEEK_CUR:
 		offset += file->f_pos;
+		break;
+	case SEEK_DATA:
+		if (offset >= i_size_read(inode)) {
+			retval = -ENXIO;
+			goto exit;
+		}
+		break;
+	case SEEK_HOLE:
+		if (offset >= i_size_read(inode)) {
+			retval = -ENXIO;
+			goto exit;
+		}
+		offset = i_size_read(inode);
+		break;
 	}
 	retval = -EINVAL;
 	if (offset >= 0 && offset <= inode->i_sb->s_maxbytes) {
diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c
index f46ae025bfb5..96a8ed91cedd 100644
--- a/fs/hpfs/dir.c
+++ b/fs/hpfs/dir.c
@@ -29,6 +29,10 @@ static loff_t hpfs_dir_lseek(struct file *filp, loff_t off, int whence)
 	struct hpfs_inode_info *hpfs_inode = hpfs_i(i);
 	struct super_block *s = i->i_sb;
 
+	/* Somebody else will have to figure out what to do here */
+	if (whence == SEEK_DATA || whence == SEEK_HOLE)
+		return -EINVAL;
+
 	hpfs_lock(s);
 
 	/*printk("dir lseek\n");*/
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 2f093ed16980..2c1705b6acd7 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -187,8 +187,11 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
 			filp->f_path.dentry->d_name.name,
 			offset, origin);
 
-	/* origin == SEEK_END => we must revalidate the cached file length */
-	if (origin == SEEK_END) {
+	/*
+	 * origin == SEEK_END || SEEK_DATA || SEEK_HOLE => we must revalidate
+	 * the cached file length
+	 */
+	if (origin != SEEK_SET || origin != SEEK_CUR) {
 		struct inode *inode = filp->f_mapping->host;
 
 		int retval = nfs_revalidate_file_size(inode, filp);
-- 
cgit v1.2.3


From 02c24a82187d5a628c68edfe71ae60dc135cd178 Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Sat, 16 Jul 2011 20:44:56 -0400
Subject: fs: push i_mutex and filemap_write_and_wait down into ->fsync()
 handlers

Btrfs needs to be able to control how filemap_write_and_wait_range() is called
in fsync to make it less of a painful operation, so push down taking i_mutex and
the calling of filemap_write_and_wait() down into the ->fsync() handlers.  Some
file systems can drop taking the i_mutex altogether it seems, like ext3 and
ocfs2.  For correctness sake I just pushed everything down in all cases to make
sure that we keep the current behavior the same for everybody, and then each
individual fs maintainer can make up their mind about what to do from there.
Thanks,

Acked-by: Jan Kara <jack@suse.cz>
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/9p/v9fs_vfs.h            |  3 ++-
 fs/9p/vfs_file.c            | 22 ++++++++++++++++++++--
 fs/affs/affs.h              |  2 +-
 fs/affs/file.c              |  8 +++++++-
 fs/afs/internal.h           |  2 +-
 fs/afs/write.c              | 18 ++++++++++++++----
 fs/bad_inode.c              |  3 ++-
 fs/block_dev.c              |  6 +-----
 fs/btrfs/ctree.h            |  2 +-
 fs/btrfs/file.c             | 21 +++++++++++++++------
 fs/ceph/caps.c              |  6 ++++--
 fs/ceph/dir.c               | 10 +++++++++-
 fs/ceph/super.h             |  3 ++-
 fs/cifs/cifsfs.h            |  4 ++--
 fs/cifs/file.c              | 18 ++++++++++++++++--
 fs/coda/coda_int.h          |  2 +-
 fs/coda/file.c              |  8 +++++++-
 fs/ecryptfs/file.c          |  7 ++++---
 fs/exofs/file.c             | 10 +++++++++-
 fs/ext2/ext2.h              |  3 ++-
 fs/ext2/file.c              |  4 ++--
 fs/ext3/fsync.c             | 18 ++++++++++++++++--
 fs/ext4/ext4.h              |  2 +-
 fs/ext4/fsync.c             | 38 +++++++++++++++++++++++++++++++++++---
 fs/fat/fat.h                |  3 ++-
 fs/fat/file.c               |  4 ++--
 fs/fuse/dir.c               |  5 +++--
 fs/fuse/file.c              | 24 ++++++++++++++++++------
 fs/fuse/fuse_i.h            |  3 ++-
 fs/gfs2/file.c              | 17 ++++++++++++++---
 fs/hfs/inode.c              |  9 ++++++++-
 fs/hfsplus/hfsplus_fs.h     |  3 ++-
 fs/hfsplus/inode.c          | 10 +++++++++-
 fs/hostfs/hostfs_kern.c     | 15 +++++++++++++--
 fs/hpfs/file.c              |  7 ++++++-
 fs/hpfs/hpfs_fn.h           |  2 +-
 fs/hppfs/hppfs.c            |  5 +++--
 fs/jffs2/file.c             |  9 ++++++++-
 fs/jffs2/os-linux.h         |  2 +-
 fs/jfs/file.c               |  9 ++++++++-
 fs/jfs/jfs_inode.h          |  2 +-
 fs/libfs.c                  | 16 ++++++++++++----
 fs/logfs/file.c             | 11 ++++++++++-
 fs/logfs/logfs.h            |  2 +-
 fs/ncpfs/file.c             |  4 ++--
 fs/nfs/dir.c                |  8 ++++++--
 fs/nfs/file.c               | 11 ++++++++---
 fs/nilfs2/file.c            | 12 ++++++++++--
 fs/nilfs2/nilfs.h           |  2 +-
 fs/ntfs/dir.c               | 10 +++++++++-
 fs/ntfs/file.c              | 10 +++++++++-
 fs/ocfs2/file.c             | 14 +++++++++++++-
 fs/reiserfs/dir.c           | 13 +++++++++++--
 fs/reiserfs/file.c          |  9 ++++++++-
 fs/sync.c                   | 25 +++----------------------
 fs/ubifs/file.c             | 21 +++++++++++----------
 fs/ubifs/ubifs.h            |  2 +-
 fs/xfs/linux-2.6/xfs_file.c | 17 ++++++++---------
 58 files changed, 398 insertions(+), 138 deletions(-)

(limited to 'fs')

diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h
index 4014160903a9..46ce357ca1ab 100644
--- a/fs/9p/v9fs_vfs.h
+++ b/fs/9p/v9fs_vfs.h
@@ -70,7 +70,8 @@ ssize_t v9fs_file_readn(struct file *, char *, char __user *, u32, u64);
 ssize_t v9fs_fid_readn(struct p9_fid *, char *, char __user *, u32, u64);
 void v9fs_blank_wstat(struct p9_wstat *wstat);
 int v9fs_vfs_setattr_dotl(struct dentry *, struct iattr *);
-int v9fs_file_fsync_dotl(struct file *filp, int datasync);
+int v9fs_file_fsync_dotl(struct file *filp, loff_t start, loff_t end,
+			 int datasync);
 ssize_t v9fs_file_write_internal(struct inode *, struct p9_fid *,
 				 const char __user *, size_t, loff_t *, int);
 int v9fs_refresh_inode(struct p9_fid *fid, struct inode *inode);
diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index ffed55817f0c..3c173fcc2c5a 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -519,32 +519,50 @@ out:
 }
 
 
-static int v9fs_file_fsync(struct file *filp, int datasync)
+static int v9fs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			   int datasync)
 {
 	struct p9_fid *fid;
+	struct inode *inode = filp->f_mapping->host;
 	struct p9_wstat wstat;
 	int retval;
 
+	retval = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (retval)
+		return retval;
+
+	mutex_lock(&inode->i_mutex);
 	P9_DPRINTK(P9_DEBUG_VFS, "filp %p datasync %x\n", filp, datasync);
 
 	fid = filp->private_data;
 	v9fs_blank_wstat(&wstat);
 
 	retval = p9_client_wstat(fid, &wstat);
+	mutex_unlock(&inode->i_mutex);
+
 	return retval;
 }
 
-int v9fs_file_fsync_dotl(struct file *filp, int datasync)
+int v9fs_file_fsync_dotl(struct file *filp, loff_t start, loff_t end,
+			 int datasync)
 {
 	struct p9_fid *fid;
+	struct inode *inode = filp->f_mapping->host;
 	int retval;
 
+	retval = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (retval)
+		return retval;
+
+	mutex_lock(&inode->i_mutex);
 	P9_DPRINTK(P9_DEBUG_VFS, "v9fs_file_fsync_dotl: filp %p datasync %x\n",
 			filp, datasync);
 
 	fid = filp->private_data;
 
 	retval = p9_client_fsync(fid, datasync);
+	mutex_unlock(&inode->i_mutex);
+
 	return retval;
 }
 
diff --git a/fs/affs/affs.h b/fs/affs/affs.h
index 0e95f73a7023..c2b9c79eb64e 100644
--- a/fs/affs/affs.h
+++ b/fs/affs/affs.h
@@ -182,7 +182,7 @@ extern int			 affs_add_entry(struct inode *dir, struct inode *inode, struct dent
 
 void		affs_free_prealloc(struct inode *inode);
 extern void	affs_truncate(struct inode *);
-int		affs_file_fsync(struct file *, int);
+int		affs_file_fsync(struct file *, loff_t, loff_t, int);
 
 /* dir.c */
 
diff --git a/fs/affs/file.c b/fs/affs/file.c
index acf321b70fcd..2f4c935cb327 100644
--- a/fs/affs/file.c
+++ b/fs/affs/file.c
@@ -923,14 +923,20 @@ affs_truncate(struct inode *inode)
 	affs_free_prealloc(inode);
 }
 
-int affs_file_fsync(struct file *filp, int datasync)
+int affs_file_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int ret, err;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	ret = write_inode_now(inode, 0);
 	err = sync_blockdev(inode->i_sb->s_bdev);
 	if (!ret)
 		ret = err;
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index f396d337b817..d2b0888126d4 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -750,7 +750,7 @@ extern void afs_pages_written_back(struct afs_vnode *, struct afs_call *);
 extern ssize_t afs_file_write(struct kiocb *, const struct iovec *,
 			      unsigned long, loff_t);
 extern int afs_writeback_all(struct afs_vnode *);
-extern int afs_fsync(struct file *, int);
+extern int afs_fsync(struct file *, loff_t, loff_t, int);
 
 
 /*****************************************************************************/
diff --git a/fs/afs/write.c b/fs/afs/write.c
index b806285ff853..9aa52d93c73c 100644
--- a/fs/afs/write.c
+++ b/fs/afs/write.c
@@ -681,9 +681,10 @@ int afs_writeback_all(struct afs_vnode *vnode)
  * - the return status from this call provides a reliable indication of
  *   whether any write errors occurred for this process.
  */
-int afs_fsync(struct file *file, int datasync)
+int afs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct dentry *dentry = file->f_path.dentry;
+	struct inode *inode = file->f_mapping->host;
 	struct afs_writeback *wb, *xwb;
 	struct afs_vnode *vnode = AFS_FS_I(dentry->d_inode);
 	int ret;
@@ -692,12 +693,19 @@ int afs_fsync(struct file *file, int datasync)
 	       vnode->fid.vid, vnode->fid.vnode, dentry->d_name.name,
 	       datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	/* use a writeback record as a marker in the queue - when this reaches
 	 * the front of the queue, all the outstanding writes are either
 	 * completed or rejected */
 	wb = kzalloc(sizeof(*wb), GFP_KERNEL);
-	if (!wb)
-		return -ENOMEM;
+	if (!wb) {
+		ret = -ENOMEM;
+		goto out;
+	}
 	wb->vnode = vnode;
 	wb->first = 0;
 	wb->last = -1;
@@ -720,7 +728,7 @@ int afs_fsync(struct file *file, int datasync)
 	if (ret < 0) {
 		afs_put_writeback(wb);
 		_leave(" = %d [wb]", ret);
-		return ret;
+		goto out;
 	}
 
 	/* wait for the preceding writes to actually complete */
@@ -729,6 +737,8 @@ int afs_fsync(struct file *file, int datasync)
 				       vnode->writebacks.next == &wb->link);
 	afs_put_writeback(wb);
 	_leave(" = %d", ret);
+out:
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/bad_inode.c b/fs/bad_inode.c
index f024d8aaddef..9205cf25f1c6 100644
--- a/fs/bad_inode.c
+++ b/fs/bad_inode.c
@@ -87,7 +87,8 @@ static int bad_file_release(struct inode *inode, struct file *filp)
 	return -EIO;
 }
 
-static int bad_file_fsync(struct file *file, int datasync)
+static int bad_file_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync)
 {
 	return -EIO;
 }
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 966617a422d9..9fb0b15331d3 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -378,7 +378,7 @@ out:
 	return retval;
 }
 	
-int blkdev_fsync(struct file *filp, int datasync)
+int blkdev_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *bd_inode = filp->f_mapping->host;
 	struct block_device *bdev = I_BDEV(bd_inode);
@@ -389,14 +389,10 @@ int blkdev_fsync(struct file *filp, int datasync)
 	 * i_mutex and doing so causes performance issues with concurrent
 	 * O_SYNC writers to a block device.
 	 */
-	mutex_unlock(&bd_inode->i_mutex);
-
 	error = blkdev_issue_flush(bdev, GFP_KERNEL, NULL);
 	if (error == -EOPNOTSUPP)
 		error = 0;
 
-	mutex_lock(&bd_inode->i_mutex);
-
 	return error;
 }
 EXPORT_SYMBOL(blkdev_fsync);
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index f1ff62bff1b3..82be74efbb26 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2605,7 +2605,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file,
 int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
 			   struct inode *inode);
 int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info);
-int btrfs_sync_file(struct file *file, int datasync);
+int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync);
 int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 			    int skip_pinned);
 extern const struct file_operations btrfs_file_operations;
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index bd4d061c6e4d..59cbdb120ad0 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1452,7 +1452,7 @@ int btrfs_release_file(struct inode *inode, struct file *filp)
  * important optimization for directories because holding the mutex prevents
  * new operations on the dir while we write to disk.
  */
-int btrfs_sync_file(struct file *file, int datasync)
+int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct dentry *dentry = file->f_path.dentry;
 	struct inode *inode = dentry->d_inode;
@@ -1462,9 +1462,13 @@ int btrfs_sync_file(struct file *file, int datasync)
 
 	trace_btrfs_sync_file(file, datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	/* we wait first, since the writeback may change the inode */
 	root->log_batch++;
-	/* the VFS called filemap_fdatawrite for us */
 	btrfs_wait_ordered_range(inode, 0, (u64)-1);
 	root->log_batch++;
 
@@ -1472,8 +1476,10 @@ int btrfs_sync_file(struct file *file, int datasync)
 	 * check the transaction that last modified this inode
 	 * and see if its already been committed
 	 */
-	if (!BTRFS_I(inode)->last_trans)
+	if (!BTRFS_I(inode)->last_trans) {
+		mutex_unlock(&inode->i_mutex);
 		goto out;
+	}
 
 	/*
 	 * if the last transaction that changed this file was before
@@ -1484,6 +1490,7 @@ int btrfs_sync_file(struct file *file, int datasync)
 	if (BTRFS_I(inode)->last_trans <=
 	    root->fs_info->last_trans_committed) {
 		BTRFS_I(inode)->last_trans = 0;
+		mutex_unlock(&inode->i_mutex);
 		goto out;
 	}
 
@@ -1496,12 +1503,15 @@ int btrfs_sync_file(struct file *file, int datasync)
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
+		mutex_unlock(&inode->i_mutex);
 		goto out;
 	}
 
 	ret = btrfs_log_dentry_safe(trans, root, dentry);
-	if (ret < 0)
+	if (ret < 0) {
+		mutex_unlock(&inode->i_mutex);
 		goto out;
+	}
 
 	/* we've logged all the items and now have a consistent
 	 * version of the file in the log.  It is possible that
@@ -1513,7 +1523,7 @@ int btrfs_sync_file(struct file *file, int datasync)
 	 * file again, but that will end up using the synchronization
 	 * inside btrfs_sync_log to keep things safe.
 	 */
-	mutex_unlock(&dentry->d_inode->i_mutex);
+	mutex_unlock(&inode->i_mutex);
 
 	if (ret != BTRFS_NO_LOG_SYNC) {
 		if (ret > 0) {
@@ -1528,7 +1538,6 @@ int btrfs_sync_file(struct file *file, int datasync)
 	} else {
 		ret = btrfs_end_transaction(trans, root);
 	}
-	mutex_lock(&dentry->d_inode->i_mutex);
 out:
 	return ret > 0 ? -EIO : ret;
 }
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index f605753c8fe9..8d74ad7ba556 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -1811,7 +1811,7 @@ out:
 	spin_unlock(&ci->i_unsafe_lock);
 }
 
-int ceph_fsync(struct file *file, int datasync)
+int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ceph_inode_info *ci = ceph_inode(inode);
@@ -1822,9 +1822,10 @@ int ceph_fsync(struct file *file, int datasync)
 	dout("fsync %p%s\n", inode, datasync ? " datasync" : "");
 	sync_write_wait(inode);
 
-	ret = filemap_write_and_wait(inode->i_mapping);
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
 	if (ret < 0)
 		return ret;
+	mutex_lock(&inode->i_mutex);
 
 	dirty = try_flush_caps(inode, NULL, &flush_tid);
 	dout("fsync dirty caps are %s\n", ceph_cap_string(dirty));
@@ -1841,6 +1842,7 @@ int ceph_fsync(struct file *file, int datasync)
 	}
 
 	dout("fsync %p%s done\n", inode, datasync ? " datasync" : "");
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 0972b457a03f..1065ac779840 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -1118,7 +1118,8 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size,
  * an fsync() on a dir will wait for any uncommitted directory
  * operations to commit.
  */
-static int ceph_dir_fsync(struct file *file, int datasync)
+static int ceph_dir_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync)
 {
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct ceph_inode_info *ci = ceph_inode(inode);
@@ -1128,6 +1129,11 @@ static int ceph_dir_fsync(struct file *file, int datasync)
 	int ret = 0;
 
 	dout("dir_fsync %p\n", inode);
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	spin_lock(&ci->i_unsafe_lock);
 	if (list_empty(head))
 		goto out;
@@ -1161,6 +1167,8 @@ static int ceph_dir_fsync(struct file *file, int datasync)
 	} while (req->r_tid < last_tid);
 out:
 	spin_unlock(&ci->i_unsafe_lock);
+	mutex_unlock(&inode->i_mutex);
+
 	return ret;
 }
 
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index 56c41ef47cad..30446b144e3d 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -728,7 +728,8 @@ extern void ceph_put_cap(struct ceph_mds_client *mdsc,
 
 extern void ceph_queue_caps_release(struct inode *inode);
 extern int ceph_write_inode(struct inode *inode, struct writeback_control *wbc);
-extern int ceph_fsync(struct file *file, int datasync);
+extern int ceph_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync);
 extern void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc,
 				    struct ceph_mds_session *session);
 extern struct ceph_cap *ceph_get_cap_for_mds(struct ceph_inode_info *ci,
diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h
index 036ca83e5f46..fbd050c8d52a 100644
--- a/fs/cifs/cifsfs.h
+++ b/fs/cifs/cifsfs.h
@@ -91,8 +91,8 @@ extern ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov,
 extern ssize_t cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov,
 				  unsigned long nr_segs, loff_t pos);
 extern int cifs_lock(struct file *, int, struct file_lock *);
-extern int cifs_fsync(struct file *, int);
-extern int cifs_strict_fsync(struct file *, int);
+extern int cifs_fsync(struct file *, loff_t, loff_t, int);
+extern int cifs_strict_fsync(struct file *, loff_t, loff_t, int);
 extern int cifs_flush(struct file *, fl_owner_t id);
 extern int cifs_file_mmap(struct file * , struct vm_area_struct *);
 extern int cifs_file_strict_mmap(struct file * , struct vm_area_struct *);
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index bb71471a4d9d..cef584451113 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -1401,7 +1401,8 @@ static int cifs_write_end(struct file *file, struct address_space *mapping,
 	return rc;
 }
 
-int cifs_strict_fsync(struct file *file, int datasync)
+int cifs_strict_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync)
 {
 	int xid;
 	int rc = 0;
@@ -1410,6 +1411,11 @@ int cifs_strict_fsync(struct file *file, int datasync)
 	struct inode *inode = file->f_path.dentry->d_inode;
 	struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb);
 
+	rc = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (rc)
+		return rc;
+	mutex_lock(&inode->i_mutex);
+
 	xid = GetXid();
 
 	cFYI(1, "Sync file - name: %s datasync: 0x%x",
@@ -1428,16 +1434,23 @@ int cifs_strict_fsync(struct file *file, int datasync)
 		rc = CIFSSMBFlush(xid, tcon, smbfile->netfid);
 
 	FreeXid(xid);
+	mutex_unlock(&inode->i_mutex);
 	return rc;
 }
 
-int cifs_fsync(struct file *file, int datasync)
+int cifs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	int xid;
 	int rc = 0;
 	struct cifs_tcon *tcon;
 	struct cifsFileInfo *smbfile = file->private_data;
 	struct cifs_sb_info *cifs_sb = CIFS_SB(file->f_path.dentry->d_sb);
+	struct inode *inode = file->f_mapping->host;
+
+	rc = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (rc)
+		return rc;
+	mutex_lock(&inode->i_mutex);
 
 	xid = GetXid();
 
@@ -1449,6 +1462,7 @@ int cifs_fsync(struct file *file, int datasync)
 		rc = CIFSSMBFlush(xid, tcon, smbfile->netfid);
 
 	FreeXid(xid);
+	mutex_unlock(&inode->i_mutex);
 	return rc;
 }
 
diff --git a/fs/coda/coda_int.h b/fs/coda/coda_int.h
index 6b443ff43a19..b7143cf783ac 100644
--- a/fs/coda/coda_int.h
+++ b/fs/coda/coda_int.h
@@ -11,7 +11,7 @@ extern int coda_fake_statfs;
 
 void coda_destroy_inodecache(void);
 int coda_init_inodecache(void);
-int coda_fsync(struct file *coda_file, int datasync);
+int coda_fsync(struct file *coda_file, loff_t start, loff_t end, int datasync);
 void coda_sysctl_init(void);
 void coda_sysctl_clean(void);
 
diff --git a/fs/coda/file.c b/fs/coda/file.c
index 0433057be330..8edd404e6419 100644
--- a/fs/coda/file.c
+++ b/fs/coda/file.c
@@ -199,7 +199,7 @@ int coda_release(struct inode *coda_inode, struct file *coda_file)
 	return 0;
 }
 
-int coda_fsync(struct file *coda_file, int datasync)
+int coda_fsync(struct file *coda_file, loff_t start, loff_t end, int datasync)
 {
 	struct file *host_file;
 	struct inode *coda_inode = coda_file->f_path.dentry->d_inode;
@@ -210,6 +210,11 @@ int coda_fsync(struct file *coda_file, int datasync)
 	      S_ISLNK(coda_inode->i_mode)))
 		return -EINVAL;
 
+	err = filemap_write_and_wait_range(coda_inode->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&coda_inode->i_mutex);
+
 	cfi = CODA_FTOC(coda_file);
 	BUG_ON(!cfi || cfi->cfi_magic != CODA_MAGIC);
 	host_file = cfi->cfi_container;
@@ -217,6 +222,7 @@ int coda_fsync(struct file *coda_file, int datasync)
 	err = vfs_fsync(host_file, datasync);
 	if (!err && !datasync)
 		err = venus_fsync(coda_inode->i_sb, coda_i2f(coda_inode));
+	mutex_unlock(&coda_inode->i_mutex);
 
 	return err;
 }
diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c
index 4ec9eb00a241..c6ac98cf9baa 100644
--- a/fs/ecryptfs/file.c
+++ b/fs/ecryptfs/file.c
@@ -270,14 +270,15 @@ static int ecryptfs_release(struct inode *inode, struct file *file)
 }
 
 static int
-ecryptfs_fsync(struct file *file, int datasync)
+ecryptfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	int rc = 0;
 
-	rc = generic_file_fsync(file, datasync);
+	rc = generic_file_fsync(file, start, end, datasync);
 	if (rc)
 		goto out;
-	rc = vfs_fsync(ecryptfs_file_to_lower(file), datasync);
+	rc = vfs_fsync_range(ecryptfs_file_to_lower(file), start, end,
+			     datasync);
 out:
 	return rc;
 }
diff --git a/fs/exofs/file.c b/fs/exofs/file.c
index 45ca323d8363..491c6c078e7f 100644
--- a/fs/exofs/file.c
+++ b/fs/exofs/file.c
@@ -42,11 +42,19 @@ static int exofs_release_file(struct inode *inode, struct file *filp)
  *   Note, in exofs all metadata is written as part of inode, regardless.
  *   The writeout is synchronous
  */
-static int exofs_file_fsync(struct file *filp, int datasync)
+static int exofs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			    int datasync)
 {
+	struct inode *inode = filp->f_mapping->host;
 	int ret;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+
+	mutex_lock(&inode->i_mutex);
 	ret = sync_inode_metadata(filp->f_mapping->host, 1);
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h
index 645be9e7ee47..af9fc89b1b2d 100644
--- a/fs/ext2/ext2.h
+++ b/fs/ext2/ext2.h
@@ -150,7 +150,8 @@ extern void ext2_write_super (struct super_block *);
 extern const struct file_operations ext2_dir_operations;
 
 /* file.c */
-extern int ext2_fsync(struct file *file, int datasync);
+extern int ext2_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync);
 extern const struct inode_operations ext2_file_inode_operations;
 extern const struct file_operations ext2_file_operations;
 extern const struct file_operations ext2_xip_file_operations;
diff --git a/fs/ext2/file.c b/fs/ext2/file.c
index 49eec9456c5b..82e06321de35 100644
--- a/fs/ext2/file.c
+++ b/fs/ext2/file.c
@@ -40,13 +40,13 @@ static int ext2_release_file (struct inode * inode, struct file * filp)
 	return 0;
 }
 
-int ext2_fsync(struct file *file, int datasync)
+int ext2_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	int ret;
 	struct super_block *sb = file->f_mapping->host->i_sb;
 	struct address_space *mapping = sb->s_bdev->bd_inode->i_mapping;
 
-	ret = generic_file_fsync(file, datasync);
+	ret = generic_file_fsync(file, start, end, datasync);
 	if (ret == -EIO || test_and_clear_bit(AS_EIO, &mapping->flags)) {
 		/* We don't really know where the IO error happened... */
 		ext2_error(sb, __func__,
diff --git a/fs/ext3/fsync.c b/fs/ext3/fsync.c
index 09b13bb34c94..0bcf63adb80a 100644
--- a/fs/ext3/fsync.c
+++ b/fs/ext3/fsync.c
@@ -43,7 +43,7 @@
  * inode to disk.
  */
 
-int ext3_sync_file(struct file *file, int datasync)
+int ext3_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ext3_inode_info *ei = EXT3_I(inode);
@@ -54,6 +54,17 @@ int ext3_sync_file(struct file *file, int datasync)
 	if (inode->i_sb->s_flags & MS_RDONLY)
 		return 0;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+
+	/*
+	 * Taking the mutex here just to keep consistent with how fsync was
+	 * called previously, however it looks like we don't need to take
+	 * i_mutex at all.
+	 */
+	mutex_lock(&inode->i_mutex);
+
 	J_ASSERT(ext3_journal_current_handle() == NULL);
 
 	/*
@@ -70,8 +81,10 @@ int ext3_sync_file(struct file *file, int datasync)
 	 *  (they were dirtied by commit).  But that's OK - the blocks are
 	 *  safe in-journal, which is all fsync() needs to ensure.
 	 */
-	if (ext3_should_journal_data(inode))
+	if (ext3_should_journal_data(inode)) {
+		mutex_unlock(&inode->i_mutex);
 		return ext3_force_commit(inode->i_sb);
+	}
 
 	if (datasync)
 		commit_tid = atomic_read(&ei->i_datasync_tid);
@@ -91,5 +104,6 @@ int ext3_sync_file(struct file *file, int datasync)
 	 */
 	if (needs_barrier)
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 1921392cd708..fa44df879711 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1758,7 +1758,7 @@ extern int ext4_htree_store_dirent(struct file *dir_file, __u32 hash,
 extern void ext4_htree_free_dir_info(struct dir_private_info *p);
 
 /* fsync.c */
-extern int ext4_sync_file(struct file *, int);
+extern int ext4_sync_file(struct file *, loff_t, loff_t, int);
 extern int ext4_flush_completed_IO(struct inode *);
 
 /* hash.c */
diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c
index ce66d2fe826c..da3bed3e0c29 100644
--- a/fs/ext4/fsync.c
+++ b/fs/ext4/fsync.c
@@ -151,6 +151,32 @@ static int ext4_sync_parent(struct inode *inode)
 	return ret;
 }
 
+/**
+ * __sync_file - generic_file_fsync without the locking and filemap_write
+ * @inode:	inode to sync
+ * @datasync:	only sync essential metadata if true
+ *
+ * This is just generic_file_fsync without the locking.  This is needed for
+ * nojournal mode to make sure this inodes data/metadata makes it to disk
+ * properly.  The i_mutex should be held already.
+ */
+static int __sync_inode(struct inode *inode, int datasync)
+{
+	int err;
+	int ret;
+
+	ret = sync_mapping_buffers(inode->i_mapping);
+	if (!(inode->i_state & I_DIRTY))
+		return ret;
+	if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
+		return ret;
+
+	err = sync_inode_metadata(inode, 1);
+	if (ret == 0)
+		ret = err;
+	return ret;
+}
+
 /*
  * akpm: A new design for ext4_sync_file().
  *
@@ -165,7 +191,7 @@ static int ext4_sync_parent(struct inode *inode)
  * i_mutex lock is held when entering and exiting this function
  */
 
-int ext4_sync_file(struct file *file, int datasync)
+int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ext4_inode_info *ei = EXT4_I(inode);
@@ -178,15 +204,20 @@ int ext4_sync_file(struct file *file, int datasync)
 
 	trace_ext4_sync_file_enter(file, datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	if (inode->i_sb->s_flags & MS_RDONLY)
-		return 0;
+		goto out;
 
 	ret = ext4_flush_completed_IO(inode);
 	if (ret < 0)
 		goto out;
 
 	if (!journal) {
-		ret = generic_file_fsync(file, datasync);
+		ret = __sync_inode(inode, datasync);
 		if (!ret && !list_empty(&inode->i_dentry))
 			ret = ext4_sync_parent(inode);
 		goto out;
@@ -220,6 +251,7 @@ int ext4_sync_file(struct file *file, int datasync)
 	if (needs_barrier)
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
  out:
+	mutex_unlock(&inode->i_mutex);
 	trace_ext4_sync_file_exit(inode, ret);
 	return ret;
 }
diff --git a/fs/fat/fat.h b/fs/fat/fat.h
index a975b4147e91..a5d3853822e0 100644
--- a/fs/fat/fat.h
+++ b/fs/fat/fat.h
@@ -310,7 +310,8 @@ extern int fat_setattr(struct dentry * dentry, struct iattr * attr);
 extern void fat_truncate_blocks(struct inode *inode, loff_t offset);
 extern int fat_getattr(struct vfsmount *mnt, struct dentry *dentry,
 		       struct kstat *stat);
-extern int fat_file_fsync(struct file *file, int datasync);
+extern int fat_file_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync);
 
 /* fat/inode.c */
 extern void fat_attach(struct inode *inode, loff_t i_pos);
diff --git a/fs/fat/file.c b/fs/fat/file.c
index e1587c54d3c1..c118acf16e43 100644
--- a/fs/fat/file.c
+++ b/fs/fat/file.c
@@ -149,12 +149,12 @@ static int fat_file_release(struct inode *inode, struct file *filp)
 	return 0;
 }
 
-int fat_file_fsync(struct file *filp, int datasync)
+int fat_file_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int res, err;
 
-	res = generic_file_fsync(filp, datasync);
+	res = generic_file_fsync(filp, start, end, datasync);
 	err = sync_mapping_buffers(MSDOS_SB(inode->i_sb)->fat_inode->i_mapping);
 
 	return res ? res : err;
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 02063dde2728..9f63e493a9b6 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -1176,9 +1176,10 @@ static int fuse_dir_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static int fuse_dir_fsync(struct file *file, int datasync)
+static int fuse_dir_fsync(struct file *file, loff_t start, loff_t end,
+			  int datasync)
 {
-	return fuse_fsync_common(file, datasync, 1);
+	return fuse_fsync_common(file, start, end, datasync, 1);
 }
 
 static bool update_mtime(unsigned ivalid)
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index 73b89df20851..7bb685cdd00c 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -400,7 +400,8 @@ static void fuse_sync_writes(struct inode *inode)
 	fuse_release_nowrite(inode);
 }
 
-int fuse_fsync_common(struct file *file, int datasync, int isdir)
+int fuse_fsync_common(struct file *file, loff_t start, loff_t end,
+		      int datasync, int isdir)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct fuse_conn *fc = get_fuse_conn(inode);
@@ -412,9 +413,15 @@ int fuse_fsync_common(struct file *file, int datasync, int isdir)
 	if (is_bad_inode(inode))
 		return -EIO;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
 	if ((!isdir && fc->no_fsync) || (isdir && fc->no_fsyncdir))
 		return 0;
 
+	mutex_lock(&inode->i_mutex);
+
 	/*
 	 * Start writeback against all dirty pages of the inode, then
 	 * wait for all outstanding writes, before sending the FSYNC
@@ -422,13 +429,15 @@ int fuse_fsync_common(struct file *file, int datasync, int isdir)
 	 */
 	err = write_inode_now(inode, 0);
 	if (err)
-		return err;
+		goto out;
 
 	fuse_sync_writes(inode);
 
 	req = fuse_get_req(fc);
-	if (IS_ERR(req))
-		return PTR_ERR(req);
+	if (IS_ERR(req)) {
+		err = PTR_ERR(req);
+		goto out;
+	}
 
 	memset(&inarg, 0, sizeof(inarg));
 	inarg.fh = ff->fh;
@@ -448,12 +457,15 @@ int fuse_fsync_common(struct file *file, int datasync, int isdir)
 			fc->no_fsync = 1;
 		err = 0;
 	}
+out:
+	mutex_unlock(&inode->i_mutex);
 	return err;
 }
 
-static int fuse_fsync(struct file *file, int datasync)
+static int fuse_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync)
 {
-	return fuse_fsync_common(file, datasync, 0);
+	return fuse_fsync_common(file, start, end, datasync, 0);
 }
 
 void fuse_read_fill(struct fuse_req *req, struct file *file, loff_t pos,
diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
index b788becada76..c6aa2d4b8517 100644
--- a/fs/fuse/fuse_i.h
+++ b/fs/fuse/fuse_i.h
@@ -589,7 +589,8 @@ void fuse_release_common(struct file *file, int opcode);
 /**
  * Send FSYNC or FSYNCDIR request
  */
-int fuse_fsync_common(struct file *file, int datasync, int isdir);
+int fuse_fsync_common(struct file *file, loff_t start, loff_t end,
+		      int datasync, int isdir);
 
 /**
  * Notify poll wakeup
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index 89c39e53760d..f82cb5e1cb6b 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -544,7 +544,9 @@ static int gfs2_close(struct inode *inode, struct file *file)
 
 /**
  * gfs2_fsync - sync the dirty data for a file (across the cluster)
- * @file: the file that points to the dentry (we ignore this)
+ * @file: the file that points to the dentry
+ * @start: the start position in the file to sync
+ * @end: the end position in the file to sync
  * @datasync: set if we can ignore timestamp changes
  *
  * The VFS will flush data for us. We only need to worry
@@ -553,23 +555,32 @@ static int gfs2_close(struct inode *inode, struct file *file)
  * Returns: errno
  */
 
-static int gfs2_fsync(struct file *file, int datasync)
+static int gfs2_fsync(struct file *file, loff_t start, loff_t end,
+		      int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	int sync_state = inode->i_state & (I_DIRTY_SYNC|I_DIRTY_DATASYNC);
 	struct gfs2_inode *ip = GFS2_I(inode);
 	int ret;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	if (datasync)
 		sync_state &= ~I_DIRTY_SYNC;
 
 	if (sync_state) {
 		ret = sync_inode_metadata(inode, 1);
-		if (ret)
+		if (ret) {
+			mutex_unlock(&inode->i_mutex);
 			return ret;
+		}
 		gfs2_ail_flush(ip->i_gl);
 	}
 
+	mutex_unlock(&inode->i_mutex);
 	return 0;
 }
 
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 5e7c3f309617..96a1b625fc74 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -627,12 +627,18 @@ int hfs_inode_setattr(struct dentry *dentry, struct iattr * attr)
 	return 0;
 }
 
-static int hfs_file_fsync(struct file *filp, int datasync)
+static int hfs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			  int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	struct super_block * sb;
 	int ret, err;
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	/* sync the inode to buffers */
 	ret = write_inode_now(inode, 0);
 
@@ -649,6 +655,7 @@ static int hfs_file_fsync(struct file *filp, int datasync)
 	err = sync_blockdev(sb->s_bdev);
 	if (!ret)
 		ret = err;
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/hfsplus/hfsplus_fs.h b/fs/hfsplus/hfsplus_fs.h
index d6857523336d..38184e360932 100644
--- a/fs/hfsplus/hfsplus_fs.h
+++ b/fs/hfsplus/hfsplus_fs.h
@@ -392,7 +392,8 @@ int hfsplus_cat_read_inode(struct inode *, struct hfs_find_data *);
 int hfsplus_cat_write_inode(struct inode *);
 struct inode *hfsplus_new_inode(struct super_block *, int);
 void hfsplus_delete_inode(struct inode *);
-int hfsplus_file_fsync(struct file *file, int datasync);
+int hfsplus_file_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync);
 
 /* ioctl.c */
 long hfsplus_ioctl(struct file *filp, unsigned int cmd, unsigned long arg);
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index 5b1cb98741cc..30486e01d003 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -308,13 +308,19 @@ static int hfsplus_setattr(struct dentry *dentry, struct iattr *attr)
 	return 0;
 }
 
-int hfsplus_file_fsync(struct file *file, int datasync)
+int hfsplus_file_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct hfsplus_inode_info *hip = HFSPLUS_I(inode);
 	struct hfsplus_sb_info *sbi = HFSPLUS_SB(inode->i_sb);
 	int error = 0, error2;
 
+	error = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (error)
+		return error;
+	mutex_lock(&inode->i_mutex);
+
 	/*
 	 * Sync inode metadata into the catalog and extent trees.
 	 */
@@ -342,6 +348,8 @@ int hfsplus_file_fsync(struct file *file, int datasync)
 	if (!test_bit(HFSPLUS_SB_NOBARRIER, &sbi->flags))
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
 
+	mutex_unlock(&inode->i_mutex);
+
 	return error;
 }
 
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index 6e449c599b9d..0d22afdd4611 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -362,9 +362,20 @@ retry:
 	return 0;
 }
 
-int hostfs_fsync(struct file *file, int datasync)
+int hostfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
-	return fsync_file(HOSTFS_I(file->f_mapping->host)->fd, datasync);
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+
+	mutex_lock(&inode->i_mutex);
+	ret = fsync_file(HOSTFS_I(inode)->fd, datasync);
+	mutex_unlock(&inode->i_mutex);
+
+	return ret;
 }
 
 static const struct file_operations hostfs_file_fops = {
diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c
index 89c500ee5213..89d2a5803ae3 100644
--- a/fs/hpfs/file.c
+++ b/fs/hpfs/file.c
@@ -18,9 +18,14 @@ static int hpfs_file_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-int hpfs_file_fsync(struct file *file, int datasync)
+int hpfs_file_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	ret = filemap_write_and_wait_range(file->f_mapping, start, end);
+	if (ret)
+		return ret;
 	return sync_blockdev(inode->i_sb->s_bdev);
 }
 
diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h
index dd552f862c8f..331b5e234ef3 100644
--- a/fs/hpfs/hpfs_fn.h
+++ b/fs/hpfs/hpfs_fn.h
@@ -258,7 +258,7 @@ void hpfs_set_ea(struct inode *, struct fnode *, const char *,
 
 /* file.c */
 
-int hpfs_file_fsync(struct file *, int);
+int hpfs_file_fsync(struct file *, loff_t, loff_t, int);
 extern const struct file_operations hpfs_file_ops;
 extern const struct inode_operations hpfs_file_iops;
 extern const struct address_space_operations hpfs_aops;
diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c
index 85c098a499f3..8635be5ffd97 100644
--- a/fs/hppfs/hppfs.c
+++ b/fs/hppfs/hppfs.c
@@ -573,9 +573,10 @@ static int hppfs_readdir(struct file *file, void *ent, filldir_t filldir)
 	return err;
 }
 
-static int hppfs_fsync(struct file *file, int datasync)
+static int hppfs_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync)
 {
-	return 0;
+	return filemap_write_and_wait_range(file->f_mapping, start, end);
 }
 
 static const struct file_operations hppfs_dir_fops = {
diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index 1c0a08d711aa..3989f7e09f7f 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -27,13 +27,20 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 			struct page **pagep, void **fsdata);
 static int jffs2_readpage (struct file *filp, struct page *pg);
 
-int jffs2_fsync(struct file *filp, int datasync)
+int jffs2_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb);
+	int ret;
+
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
 
+	mutex_lock(&inode->i_mutex);
 	/* Trigger GC to flush any pending writes for this inode */
 	jffs2_flush_wbuf_gc(c, inode->i_ino);
+	mutex_unlock(&inode->i_mutex);
 
 	return 0;
 }
diff --git a/fs/jffs2/os-linux.h b/fs/jffs2/os-linux.h
index 65c6c43ca482..9c252835e8e5 100644
--- a/fs/jffs2/os-linux.h
+++ b/fs/jffs2/os-linux.h
@@ -158,7 +158,7 @@ extern const struct inode_operations jffs2_dir_inode_operations;
 extern const struct file_operations jffs2_file_operations;
 extern const struct inode_operations jffs2_file_inode_operations;
 extern const struct address_space_operations jffs2_file_address_operations;
-int jffs2_fsync(struct file *, int);
+int jffs2_fsync(struct file *, loff_t, loff_t, int);
 int jffs2_do_readpage_unlock (struct inode *inode, struct page *pg);
 
 /* ioctl.c */
diff --git a/fs/jfs/file.c b/fs/jfs/file.c
index 9f32315acef1..7527855b5cc6 100644
--- a/fs/jfs/file.c
+++ b/fs/jfs/file.c
@@ -28,19 +28,26 @@
 #include "jfs_acl.h"
 #include "jfs_debug.h"
 
-int jfs_fsync(struct file *file, int datasync)
+int jfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	int rc = 0;
 
+	rc = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (rc)
+		return rc;
+
+	mutex_lock(&inode->i_mutex);
 	if (!(inode->i_state & I_DIRTY) ||
 	    (datasync && !(inode->i_state & I_DIRTY_DATASYNC))) {
 		/* Make sure committed changes hit the disk */
 		jfs_flush_journal(JFS_SBI(inode->i_sb)->log, 1);
+		mutex_unlock(&inode->i_mutex);
 		return rc;
 	}
 
 	rc |= jfs_commit_inode(inode, 1);
+	mutex_unlock(&inode->i_mutex);
 
 	return rc ? -EIO : 0;
 }
diff --git a/fs/jfs/jfs_inode.h b/fs/jfs/jfs_inode.h
index ec2fb8b945fc..9271cfe4a149 100644
--- a/fs/jfs/jfs_inode.h
+++ b/fs/jfs/jfs_inode.h
@@ -21,7 +21,7 @@
 struct fid;
 
 extern struct inode *ialloc(struct inode *, umode_t);
-extern int jfs_fsync(struct file *, int);
+extern int jfs_fsync(struct file *, loff_t, loff_t, int);
 extern long jfs_ioctl(struct file *, unsigned int, unsigned long);
 extern long jfs_compat_ioctl(struct file *, unsigned int, unsigned long);
 extern struct inode *jfs_iget(struct super_block *, unsigned long);
diff --git a/fs/libfs.c b/fs/libfs.c
index bd50b11f92da..8f2271a5df53 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -905,21 +905,29 @@ EXPORT_SYMBOL_GPL(generic_fh_to_parent);
  * filesystems which track all non-inode metadata in the buffers list
  * hanging off the address_space structure.
  */
-int generic_file_fsync(struct file *file, int datasync)
+int generic_file_fsync(struct file *file, loff_t start, loff_t end,
+		       int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	int err;
 	int ret;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	ret = sync_mapping_buffers(inode->i_mapping);
 	if (!(inode->i_state & I_DIRTY))
-		return ret;
+		goto out;
 	if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
-		return ret;
+		goto out;
 
 	err = sync_inode_metadata(inode, 1);
 	if (ret == 0)
 		ret = err;
+out:
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 EXPORT_SYMBOL(generic_file_fsync);
@@ -956,7 +964,7 @@ EXPORT_SYMBOL(generic_check_addressable);
 /*
  * No-op implementation of ->fsync for in-memory filesystems.
  */
-int noop_fsync(struct file *file, int datasync)
+int noop_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	return 0;
 }
diff --git a/fs/logfs/file.c b/fs/logfs/file.c
index c2ad7028def4..b548c87a86f1 100644
--- a/fs/logfs/file.c
+++ b/fs/logfs/file.c
@@ -219,11 +219,20 @@ long logfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	}
 }
 
-int logfs_fsync(struct file *file, int datasync)
+int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct super_block *sb = file->f_mapping->host->i_sb;
+	struct inode *inode = file->f_mapping->host;
+	int ret;
+
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
 
+	mutex_lock(&inode->i_mutex);
 	logfs_write_anchor(sb);
+	mutex_unlock(&inode->i_mutex);
+
 	return 0;
 }
 
diff --git a/fs/logfs/logfs.h b/fs/logfs/logfs.h
index 57afd4a6fabb..f22d108bfa5d 100644
--- a/fs/logfs/logfs.h
+++ b/fs/logfs/logfs.h
@@ -506,7 +506,7 @@ extern const struct file_operations logfs_reg_fops;
 extern const struct address_space_operations logfs_reg_aops;
 int logfs_readpage(struct file *file, struct page *page);
 long logfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
-int logfs_fsync(struct file *file, int datasync);
+int logfs_fsync(struct file *file, loff_t start, loff_t end, int datasync);
 
 /* gc.c */
 u32 get_best_cand(struct super_block *sb, struct candidate_list *list, u32 *ec);
diff --git a/fs/ncpfs/file.c b/fs/ncpfs/file.c
index 0ed65e0c3dfe..64a326418aa2 100644
--- a/fs/ncpfs/file.c
+++ b/fs/ncpfs/file.c
@@ -20,9 +20,9 @@
 
 #include "ncp_fs.h"
 
-static int ncp_fsync(struct file *file, int datasync)
+static int ncp_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
-	return 0;
+	return filemap_write_and_wait_range(file->f_mapping, start, end);
 }
 
 /*
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 8a45e6d1f6a4..57f578e2560a 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -56,7 +56,7 @@ static int nfs_link(struct dentry *, struct inode *, struct dentry *);
 static int nfs_mknod(struct inode *, struct dentry *, int, dev_t);
 static int nfs_rename(struct inode *, struct dentry *,
 		      struct inode *, struct dentry *);
-static int nfs_fsync_dir(struct file *, int);
+static int nfs_fsync_dir(struct file *, loff_t, loff_t, int);
 static loff_t nfs_llseek_dir(struct file *, loff_t, int);
 static void nfs_readdir_clear_array(struct page*);
 
@@ -945,15 +945,19 @@ out:
  * All directory operations under NFS are synchronous, so fsync()
  * is a dummy operation.
  */
-static int nfs_fsync_dir(struct file *filp, int datasync)
+static int nfs_fsync_dir(struct file *filp, loff_t start, loff_t end,
+			 int datasync)
 {
 	struct dentry *dentry = filp->f_path.dentry;
+	struct inode *inode = dentry->d_inode;
 
 	dfprintk(FILE, "NFS: fsync dir(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
+	mutex_lock(&inode->i_mutex);
 	nfs_inc_stats(dentry->d_inode, NFSIOS_VFSFSYNC);
+	mutex_unlock(&inode->i_mutex);
 	return 0;
 }
 
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 2c1705b6acd7..28b8c3f3cda3 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -55,7 +55,7 @@ static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe,
 static ssize_t nfs_file_write(struct kiocb *, const struct iovec *iov,
 				unsigned long nr_segs, loff_t pos);
 static int  nfs_file_flush(struct file *, fl_owner_t id);
-static int  nfs_file_fsync(struct file *, int datasync);
+static int  nfs_file_fsync(struct file *, loff_t, loff_t, int datasync);
 static int nfs_check_flags(int flags);
 static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl);
 static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl);
@@ -308,7 +308,7 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma)
  * fall back to doing a synchronous write.
  */
 static int
-nfs_file_fsync(struct file *file, int datasync)
+nfs_file_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct dentry *dentry = file->f_path.dentry;
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
@@ -316,11 +316,15 @@ nfs_file_fsync(struct file *file, int datasync)
 	int have_error, status;
 	int ret = 0;
 
-
 	dprintk("NFS: fsync file(%s/%s) datasync %d\n",
 			dentry->d_parent->d_name.name, dentry->d_name.name,
 			datasync);
 
+	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (ret)
+		return ret;
+	mutex_lock(&inode->i_mutex);
+
 	nfs_inc_stats(inode, NFSIOS_VFSFSYNC);
 	have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
 	status = nfs_commit_inode(inode, FLUSH_SYNC);
@@ -332,6 +336,7 @@ nfs_file_fsync(struct file *file, int datasync)
 	if (!ret && !datasync)
 		/* application has asked for meta-data sync */
 		ret = pnfs_layoutcommit_inode(inode, true);
+	mutex_unlock(&inode->i_mutex);
 	return ret;
 }
 
diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c
index d7eeca62febd..26601529dc17 100644
--- a/fs/nilfs2/file.c
+++ b/fs/nilfs2/file.c
@@ -27,7 +27,7 @@
 #include "nilfs.h"
 #include "segment.h"
 
-int nilfs_sync_file(struct file *file, int datasync)
+int nilfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	/*
 	 * Called from fsync() system call
@@ -40,8 +40,15 @@ int nilfs_sync_file(struct file *file, int datasync)
 	struct inode *inode = file->f_mapping->host;
 	int err;
 
-	if (!nilfs_inode_dirty(inode))
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&inode->i_mutex);
+
+	if (!nilfs_inode_dirty(inode)) {
+		mutex_unlock(&inode->i_mutex);
 		return 0;
+	}
 
 	if (datasync)
 		err = nilfs_construct_dsync_segment(inode->i_sb, inode, 0,
@@ -49,6 +56,7 @@ int nilfs_sync_file(struct file *file, int datasync)
 	else
 		err = nilfs_construct_segment(inode->i_sb);
 
+	mutex_unlock(&inode->i_mutex);
 	return err;
 }
 
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index 6fb7511c0328..255d5e1c03b7 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -235,7 +235,7 @@ extern void nilfs_set_link(struct inode *, struct nilfs_dir_entry *,
 			   struct page *, struct inode *);
 
 /* file.c */
-extern int nilfs_sync_file(struct file *, int);
+extern int nilfs_sync_file(struct file *, loff_t, loff_t, int);
 
 /* ioctl.c */
 long nilfs_ioctl(struct file *, unsigned int, unsigned long);
diff --git a/fs/ntfs/dir.c b/fs/ntfs/dir.c
index 0f48e7c5d9e1..99e36107ff60 100644
--- a/fs/ntfs/dir.c
+++ b/fs/ntfs/dir.c
@@ -1527,13 +1527,20 @@ static int ntfs_dir_open(struct inode *vi, struct file *filp)
  * this problem for now.  We do write the $BITMAP attribute if it is present
  * which is the important one for a directory so things are not too bad.
  */
-static int ntfs_dir_fsync(struct file *filp, int datasync)
+static int ntfs_dir_fsync(struct file *filp, loff_t start, loff_t end,
+			  int datasync)
 {
 	struct inode *bmp_vi, *vi = filp->f_mapping->host;
 	int err, ret;
 	ntfs_attr na;
 
 	ntfs_debug("Entering for inode 0x%lx.", vi->i_ino);
+
+	err = filemap_write_and_wait_range(vi->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&vi->i_mutex);
+
 	BUG_ON(!S_ISDIR(vi->i_mode));
 	/* If the bitmap attribute inode is in memory sync it, too. */
 	na.mft_no = vi->i_ino;
@@ -1555,6 +1562,7 @@ static int ntfs_dir_fsync(struct file *filp, int datasync)
 	else
 		ntfs_warning(vi->i_sb, "Failed to f%ssync inode 0x%lx.  Error "
 				"%u.", datasync ? "data" : "", vi->i_ino, -ret);
+	mutex_unlock(&vi->i_mutex);
 	return ret;
 }
 
diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c
index b59f5ac26bef..c587e2d27183 100644
--- a/fs/ntfs/file.c
+++ b/fs/ntfs/file.c
@@ -2152,12 +2152,19 @@ static ssize_t ntfs_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
  * with this inode but since we have no simple way of getting to them we ignore
  * this problem for now.
  */
-static int ntfs_file_fsync(struct file *filp, int datasync)
+static int ntfs_file_fsync(struct file *filp, loff_t start, loff_t end,
+			   int datasync)
 {
 	struct inode *vi = filp->f_mapping->host;
 	int err, ret = 0;
 
 	ntfs_debug("Entering for inode 0x%lx.", vi->i_ino);
+
+	err = filemap_write_and_wait_range(vi->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&vi->i_mutex);
+
 	BUG_ON(S_ISDIR(vi->i_mode));
 	if (!datasync || !NInoNonResident(NTFS_I(vi)))
 		ret = __ntfs_write_inode(vi, 1);
@@ -2175,6 +2182,7 @@ static int ntfs_file_fsync(struct file *filp, int datasync)
 	else
 		ntfs_warning(vi->i_sb, "Failed to f%ssync inode 0x%lx.  Error "
 				"%u.", datasync ? "data" : "", vi->i_ino, -ret);
+	mutex_unlock(&vi->i_mutex);
 	return ret;
 }
 
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 22d604601957..0fc2bd34039d 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -171,7 +171,8 @@ static int ocfs2_dir_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static int ocfs2_sync_file(struct file *file, int datasync)
+static int ocfs2_sync_file(struct file *file, loff_t start, loff_t end,
+			   int datasync)
 {
 	int err = 0;
 	journal_t *journal;
@@ -184,6 +185,16 @@ static int ocfs2_sync_file(struct file *file, int datasync)
 			      file->f_path.dentry->d_name.name,
 			      (unsigned long long)datasync);
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	/*
+	 * Probably don't need the i_mutex at all in here, just putting it here
+	 * to be consistent with how fsync used to be called, someone more
+	 * familiar with the fs could possibly remove it.
+	 */
+	mutex_lock(&inode->i_mutex);
 	if (datasync && !(inode->i_state & I_DIRTY_DATASYNC)) {
 		/*
 		 * We still have to flush drive's caches to get data to the
@@ -200,6 +211,7 @@ static int ocfs2_sync_file(struct file *file, int datasync)
 bail:
 	if (err)
 		mlog_errno(err);
+	mutex_unlock(&inode->i_mutex);
 
 	return (err < 0) ? -EIO : 0;
 }
diff --git a/fs/reiserfs/dir.c b/fs/reiserfs/dir.c
index 198dabf1b2bb..133e9355dc6f 100644
--- a/fs/reiserfs/dir.c
+++ b/fs/reiserfs/dir.c
@@ -14,7 +14,8 @@
 extern const struct reiserfs_key MIN_KEY;
 
 static int reiserfs_readdir(struct file *, void *, filldir_t);
-static int reiserfs_dir_fsync(struct file *filp, int datasync);
+static int reiserfs_dir_fsync(struct file *filp, loff_t start, loff_t end,
+			      int datasync);
 
 const struct file_operations reiserfs_dir_operations = {
 	.llseek = generic_file_llseek,
@@ -27,13 +28,21 @@ const struct file_operations reiserfs_dir_operations = {
 #endif
 };
 
-static int reiserfs_dir_fsync(struct file *filp, int datasync)
+static int reiserfs_dir_fsync(struct file *filp, loff_t start, loff_t end,
+			      int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int err;
+
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	reiserfs_write_lock(inode->i_sb);
 	err = reiserfs_commit_for_inode(inode);
 	reiserfs_write_unlock(inode->i_sb);
+	mutex_unlock(&inode->i_mutex);
 	if (err < 0)
 		return err;
 	return 0;
diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c
index bbf31003d308..c7156dc39ce7 100644
--- a/fs/reiserfs/file.c
+++ b/fs/reiserfs/file.c
@@ -140,12 +140,18 @@ static void reiserfs_vfs_truncate_file(struct inode *inode)
  * be removed...
  */
 
-static int reiserfs_sync_file(struct file *filp, int datasync)
+static int reiserfs_sync_file(struct file *filp, loff_t start, loff_t end,
+			      int datasync)
 {
 	struct inode *inode = filp->f_mapping->host;
 	int err;
 	int barrier_done;
 
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+
+	mutex_lock(&inode->i_mutex);
 	BUG_ON(!S_ISREG(inode->i_mode));
 	err = sync_mapping_buffers(inode->i_mapping);
 	reiserfs_write_lock(inode->i_sb);
@@ -153,6 +159,7 @@ static int reiserfs_sync_file(struct file *filp, int datasync)
 	reiserfs_write_unlock(inode->i_sb);
 	if (barrier_done != 1 && reiserfs_barrier_flush(inode->i_sb))
 		blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
+	mutex_unlock(&inode->i_mutex);
 	if (barrier_done < 0)
 		return barrier_done;
 	return (err < 0) ? -EIO : 0;
diff --git a/fs/sync.c b/fs/sync.c
index c38ec163da6c..c98a7477edfd 100644
--- a/fs/sync.c
+++ b/fs/sync.c
@@ -165,28 +165,9 @@ SYSCALL_DEFINE1(syncfs, int, fd)
  */
 int vfs_fsync_range(struct file *file, loff_t start, loff_t end, int datasync)
 {
-	struct address_space *mapping = file->f_mapping;
-	int err, ret;
-
-	if (!file->f_op || !file->f_op->fsync) {
-		ret = -EINVAL;
-		goto out;
-	}
-
-	ret = filemap_write_and_wait_range(mapping, start, end);
-
-	/*
-	 * We need to protect against concurrent writers, which could cause
-	 * livelocks in fsync_buffers_list().
-	 */
-	mutex_lock(&mapping->host->i_mutex);
-	err = file->f_op->fsync(file, datasync);
-	if (!ret)
-		ret = err;
-	mutex_unlock(&mapping->host->i_mutex);
-
-out:
-	return ret;
+	if (!file->f_op || !file->f_op->fsync)
+		return -EINVAL;
+	return file->f_op->fsync(file, start, end, datasync);
 }
 EXPORT_SYMBOL(vfs_fsync_range);
 
diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 5e7fccfc4b29..89ef9a2f7837 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -1304,7 +1304,7 @@ static void *ubifs_follow_link(struct dentry *dentry, struct nameidata *nd)
 	return NULL;
 }
 
-int ubifs_fsync(struct file *file, int datasync)
+int ubifs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
 	struct inode *inode = file->f_mapping->host;
 	struct ubifs_info *c = inode->i_sb->s_fs_info;
@@ -1319,14 +1319,16 @@ int ubifs_fsync(struct file *file, int datasync)
 		 */
 		return 0;
 
-	/*
-	 * VFS has already synchronized dirty pages for this inode. Synchronize
-	 * the inode unless this is a 'datasync()' call.
-	 */
+	err = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (err)
+		return err;
+	mutex_lock(&inode->i_mutex);
+
+	/* Synchronize the inode unless this is a 'datasync()' call. */
 	if (!datasync || (inode->i_state & I_DIRTY_DATASYNC)) {
 		err = inode->i_sb->s_op->write_inode(inode, NULL);
 		if (err)
-			return err;
+			goto out;
 	}
 
 	/*
@@ -1334,10 +1336,9 @@ int ubifs_fsync(struct file *file, int datasync)
 	 * them.
 	 */
 	err = ubifs_sync_wbufs_by_inode(c, inode);
-	if (err)
-		return err;
-
-	return 0;
+out:
+	mutex_unlock(&inode->i_mutex);
+	return err;
 }
 
 /**
diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h
index f79983d6f860..4cd648501fa4 100644
--- a/fs/ubifs/ubifs.h
+++ b/fs/ubifs/ubifs.h
@@ -1720,7 +1720,7 @@ const struct ubifs_lprops *ubifs_fast_find_frdi_idx(struct ubifs_info *c);
 int ubifs_calc_dark(const struct ubifs_info *c, int spc);
 
 /* file.c */
-int ubifs_fsync(struct file *file, int datasync);
+int ubifs_fsync(struct file *file, loff_t start, loff_t end, int datasync);
 int ubifs_setattr(struct dentry *dentry, struct iattr *attr);
 
 /* dir.c */
diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c
index 7f782af286bf..fbbf657df0cd 100644
--- a/fs/xfs/linux-2.6/xfs_file.c
+++ b/fs/xfs/linux-2.6/xfs_file.c
@@ -127,6 +127,8 @@ xfs_iozero(
 STATIC int
 xfs_file_fsync(
 	struct file		*file,
+	loff_t			start,
+	loff_t			end,
 	int			datasync)
 {
 	struct inode		*inode = file->f_mapping->host;
@@ -138,6 +140,10 @@ xfs_file_fsync(
 
 	trace_xfs_file_fsync(ip);
 
+	error = filemap_write_and_wait_range(inode->i_mapping, start, end);
+	if (error)
+		return error;
+
 	if (XFS_FORCED_SHUTDOWN(mp))
 		return -XFS_ERROR(EIO);
 
@@ -875,18 +881,11 @@ xfs_file_aio_write(
 	/* Handle various SYNC-type writes */
 	if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) {
 		loff_t end = pos + ret - 1;
-		int error, error2;
 
 		xfs_rw_iunlock(ip, iolock);
-		error = filemap_write_and_wait_range(mapping, pos, end);
+		ret = -xfs_file_fsync(file, pos, end,
+				      (file->f_flags & __O_SYNC) ? 0 : 1);
 		xfs_rw_ilock(ip, iolock);
-
-		error2 = -xfs_file_fsync(file,
-					 (file->f_flags & __O_SYNC) ? 0 : 1);
-		if (error)
-			ret = error;
-		else if (error2)
-			ret = error2;
 	}
 
 out_unlock:
-- 
cgit v1.2.3


From 2fbe8c8ad176db69dfc682b4abb721c8fc19c3fd Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 16 Jul 2011 21:38:06 -0400
Subject: get rid of useless dget_parent() in fs/btrfs/ioctl.c

both callers there have dentry->d_parent stabilized by the fact that
their caller had obtained dentry from lookup_one_len() and had not
dropped ->i_mutex on parent since then.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/btrfs/ioctl.c | 16 ++++------------
 1 file changed, 4 insertions(+), 12 deletions(-)

(limited to 'fs')

diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index a3c4751e07db..622543309eb2 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -323,7 +323,7 @@ static noinline int create_subvol(struct btrfs_root *root,
 	struct btrfs_inode_item *inode_item;
 	struct extent_buffer *leaf;
 	struct btrfs_root *new_root;
-	struct dentry *parent = dget_parent(dentry);
+	struct dentry *parent = dentry->d_parent;
 	struct inode *dir;
 	int ret;
 	int err;
@@ -332,10 +332,8 @@ static noinline int create_subvol(struct btrfs_root *root,
 	u64 index = 0;
 
 	ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
-	if (ret) {
-		dput(parent);
+	if (ret)
 		return ret;
-	}
 
 	dir = parent->d_inode;
 
@@ -346,10 +344,8 @@ static noinline int create_subvol(struct btrfs_root *root,
 	 * 2 - dir items
 	 */
 	trans = btrfs_start_transaction(root, 6);
-	if (IS_ERR(trans)) {
-		dput(parent);
+	if (IS_ERR(trans))
 		return PTR_ERR(trans);
-	}
 
 	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
 				      0, objectid, NULL, 0, 0, 0);
@@ -439,7 +435,6 @@ static noinline int create_subvol(struct btrfs_root *root,
 
 	d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry));
 fail:
-	dput(parent);
 	if (async_transid) {
 		*async_transid = trans->transid;
 		err = btrfs_commit_transaction_async(trans, root, 1);
@@ -456,7 +451,6 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
 			   bool readonly)
 {
 	struct inode *inode;
-	struct dentry *parent;
 	struct btrfs_pending_snapshot *pending_snapshot;
 	struct btrfs_trans_handle *trans;
 	int ret;
@@ -504,9 +498,7 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
 	if (ret)
 		goto fail;
 
-	parent = dget_parent(dentry);
-	inode = btrfs_lookup_dentry(parent->d_inode, dentry);
-	dput(parent);
+	inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry);
 	if (IS_ERR(inode)) {
 		ret = PTR_ERR(inode);
 		goto fail;
-- 
cgit v1.2.3


From 10d9f309d88ca7f47133d57e99b72810f119f75b Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sat, 16 Jul 2011 23:09:10 -0400
Subject: get rid of useless dget_parent() in btrfs rename() and link()

->d_parent is locked and stable there...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/btrfs/inode.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

(limited to 'fs')

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index bcb20a9a3b93..2548a04a0230 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -4766,11 +4766,10 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 	if (err) {
 		drop_inode = 1;
 	} else {
-		struct dentry *parent = dget_parent(dentry);
+		struct dentry *parent = dentry->d_parent;
 		err = btrfs_update_inode(trans, root, inode);
 		BUG_ON(err);
 		btrfs_log_new_name(trans, inode, NULL, parent);
-		dput(parent);
 	}
 
 	nr = trans->blocks_used;
@@ -7062,9 +7061,8 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	BUG_ON(ret);
 
 	if (old_ino != BTRFS_FIRST_FREE_OBJECTID) {
-		struct dentry *parent = dget_parent(new_dentry);
+		struct dentry *parent = new_dentry->d_parent;
 		btrfs_log_new_name(trans, old_inode, old_dir, parent);
-		dput(parent);
 		btrfs_end_log_trans(root);
 	}
 out_fail:
-- 
cgit v1.2.3


From 79ac5a46c5c1c17476fbf84b4d4600d6d565defd Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 17 Jul 2011 10:17:46 -0400
Subject: jfs_lookup(): don't bother with . or ..

they'll never be passed to ->lookup()

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/jfs/namei.c | 39 +++++++++++++++------------------------
 1 file changed, 15 insertions(+), 24 deletions(-)

(limited to 'fs')

diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c
index 247331551992..03787ef6a118 100644
--- a/fs/jfs/namei.c
+++ b/fs/jfs/namei.c
@@ -1456,34 +1456,25 @@ static struct dentry *jfs_lookup(struct inode *dip, struct dentry *dentry, struc
 	ino_t inum;
 	struct inode *ip;
 	struct component_name key;
-	const char *name = dentry->d_name.name;
-	int len = dentry->d_name.len;
 	int rc;
 
-	jfs_info("jfs_lookup: name = %s", name);
-
-	if ((name[0] == '.') && (len == 1))
-		inum = dip->i_ino;
-	else if (strcmp(name, "..") == 0)
-		inum = PARENT(dip);
-	else {
-		if ((rc = get_UCSname(&key, dentry)))
-			return ERR_PTR(rc);
-		rc = dtSearch(dip, &key, &inum, &btstack, JFS_LOOKUP);
-		free_UCSname(&key);
-		if (rc == -ENOENT) {
-			d_add(dentry, NULL);
-			return NULL;
-		} else if (rc) {
-			jfs_err("jfs_lookup: dtSearch returned %d", rc);
-			return ERR_PTR(rc);
-		}
+	jfs_info("jfs_lookup: name = %s", dentry->d_name.name);
+
+	if ((rc = get_UCSname(&key, dentry)))
+		return ERR_PTR(rc);
+	rc = dtSearch(dip, &key, &inum, &btstack, JFS_LOOKUP);
+	free_UCSname(&key);
+	if (rc == -ENOENT) {
+		ip = NULL;
+	} else if (rc) {
+		jfs_err("jfs_lookup: dtSearch returned %d", rc);
+		ip = ERR_PTR(rc);
+	} else {
+		ip = jfs_iget(dip->i_sb, inum);
+		if (IS_ERR(ip))
+			jfs_err("jfs_lookup: iget failed on inum %d", (uint)inum);
 	}
 
-	ip = jfs_iget(dip->i_sb, inum);
-	if (IS_ERR(ip))
-		jfs_err("jfs_lookup: iget failed on inum %d", (uint) inum);
-
 	return d_splice_alias(ip, dentry);
 }
 
-- 
cgit v1.2.3


From 6c673ab393bc18e8bff729cd04cf384d15e72a04 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 17 Jul 2011 10:22:27 -0400
Subject: simplify gfs2_lookup()

d_splice_alias() will DTRT when given NULL or ERR_PTR

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/gfs2/inode.c | 14 +++-----------
 1 file changed, 3 insertions(+), 11 deletions(-)

(limited to 'fs')

diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index b1090d66a6fd..0fb51a96eff0 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -792,13 +792,8 @@ static int gfs2_create(struct inode *dir, struct dentry *dentry,
 static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry,
 				  struct nameidata *nd)
 {
-	struct inode *inode = NULL;
-
-	inode = gfs2_lookupi(dir, &dentry->d_name, 0);
-	if (inode && IS_ERR(inode))
-		return ERR_CAST(inode);
-
-	if (inode) {
+	struct inode *inode = gfs2_lookupi(dir, &dentry->d_name, 0);
+	if (inode && !IS_ERR(inode)) {
 		struct gfs2_glock *gl = GFS2_I(inode)->i_gl;
 		struct gfs2_holder gh;
 		int error;
@@ -808,11 +803,8 @@ static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry,
 			return ERR_PTR(error);
 		}
 		gfs2_glock_dq_uninit(&gh);
-		return d_splice_alias(inode, dentry);
 	}
-	d_add(dentry, inode);
-
-	return NULL;
+	return d_splice_alias(inode, dentry);
 }
 
 /**
-- 
cgit v1.2.3


From 4513d899c418ff69052420e29e354e4c64b3ef76 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Sun, 17 Jul 2011 10:52:14 -0400
Subject: switch d_add_ci() to d_splice_alias() in "found negative" case as
 well

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c | 24 +++++-------------------
 1 file changed, 5 insertions(+), 19 deletions(-)

(limited to 'fs')

diff --git a/fs/dcache.c b/fs/dcache.c
index 2762804a140d..d1d6b3349ec7 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -1652,26 +1652,12 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode,
 	 * Negative dentry: instantiate it unless the inode is a directory and
 	 * already has a dentry.
 	 */
-	spin_lock(&inode->i_lock);
-	if (!S_ISDIR(inode->i_mode) || list_empty(&inode->i_dentry)) {
-		__d_instantiate(found, inode);
-		spin_unlock(&inode->i_lock);
-		security_d_instantiate(found, inode);
-		return found;
+	new = d_splice_alias(inode, found);
+	if (new) {
+		dput(found);
+		found = new;
 	}
-
-	/*
-	 * In case a directory already has a (disconnected) entry grab a
-	 * reference to it, move it in place and use it.
-	 */
-	new = list_entry(inode->i_dentry.next, struct dentry, d_alias);
-	__dget(new);
-	spin_unlock(&inode->i_lock);
-	security_d_instantiate(found, inode);
-	d_move(new, found);
-	iput(inode);
-	dput(found);
-	return new;
+	return found;
 
 err_out:
 	iput(inode);
-- 
cgit v1.2.3


From e4b9f0058145a3bed4199eacee62a9969c163401 Mon Sep 17 00:00:00 2001
From: David Howells <dhowells@redhat.com>
Date: Mon, 18 Jul 2011 13:50:00 +0100
Subject: AFS: Fix silly characters in a comment

Fix silly characters in a comment in AFS code (some weird characters replaced
the word 'flag' some point way back).

Reported-by: viro@ZenIV.linux.org.uk
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/afs/afs_vl.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/afs/afs_vl.h b/fs/afs/afs_vl.h
index 8bbefe009ed4..800f607ffaf5 100644
--- a/fs/afs/afs_vl.h
+++ b/fs/afs/afs_vl.h
@@ -49,7 +49,7 @@ enum AFSVL_Errors {
 	AFSVL_BADVOLOPER 	= 363542,	/* Bad volume operation code */
 	AFSVL_BADRELLOCKTYPE 	= 363543,	/* Bad release lock type */
 	AFSVL_RERELEASE 	= 363544,	/* Status report: last release was aborted */
-	AFSVL_BADSERVERFLAG 	= 363545,	/* Invalid replication site server °ag */
+	AFSVL_BADSERVERFLAG 	= 363545,	/* Invalid replication site server flag */
 	AFSVL_PERM 		= 363546,	/* No permission access */
 	AFSVL_NOMEM 		= 363547,	/* malloc/realloc failed to alloc enough memory */
 };
-- 
cgit v1.2.3


From 86c98e8cdb21ff4628f4d48559ab6e006380fa4b Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Mon, 18 Jul 2011 23:39:07 -0400
Subject: Remove dead code in dget_parent()

->d_parent is never NULL...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/dcache.c | 5 -----
 1 file changed, 5 deletions(-)

(limited to 'fs')

diff --git a/fs/dcache.c b/fs/dcache.c
index d1d6b3349ec7..3c34ac0e9a1b 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -549,10 +549,6 @@ repeat:
 	 */
 	rcu_read_lock();
 	ret = dentry->d_parent;
-	if (!ret) {
-		rcu_read_unlock();
-		goto out;
-	}
 	spin_lock(&ret->d_lock);
 	if (unlikely(ret != dentry->d_parent)) {
 		spin_unlock(&ret->d_lock);
@@ -563,7 +559,6 @@ repeat:
 	BUG_ON(!ret->d_count);
 	ret->d_count++;
 	spin_unlock(&ret->d_lock);
-out:
 	return ret;
 }
 EXPORT_SYMBOL(dget_parent);
-- 
cgit v1.2.3


From 841590ce16c19a3ce38028adfc8b1955482ee00c Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 21 Jul 2011 15:49:09 -0400
Subject: fix IN_DELETE_SELF on overwriting rename() on ramfs et.al.

On ramfs and other simple_rename() users IN_DELETE_SELF is not generated
for victim of overwriting rename() if it's is a directory.  Works on
most of the local filesystems and really trivial to fix...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/libfs.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/libfs.c b/fs/libfs.c
index 8f2271a5df53..c18e9a1235b6 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -328,8 +328,10 @@ int simple_rename(struct inode *old_dir, struct dentry *old_dentry,
 
 	if (new_dentry->d_inode) {
 		simple_unlink(new_dir, new_dentry);
-		if (they_are_dirs)
+		if (they_are_dirs) {
+			drop_nlink(new_dentry->d_inode);
 			drop_nlink(old_dir);
+		}
 	} else if (they_are_dirs) {
 		drop_nlink(old_dir);
 		inc_nlink(new_dir);
-- 
cgit v1.2.3


From 22ba747f660c0acd14761628c24aa972d18058a0 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 21 Jul 2011 15:57:47 -0400
Subject: jffs2: fix IN_DELETE_SELF on overwriting rename() killing a directory

We don't generate IN_DELETE_SELF on victim of overwriting rename() if
it happens to be a directory.  Trivially fixed by doing to ->i_nlink
what we do ->pino_nlink a couple of lines later in jffs2_rename().

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/jffs2/dir.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c
index 8f40ce4f1777..5f243cd63afc 100644
--- a/fs/jffs2/dir.c
+++ b/fs/jffs2/dir.c
@@ -820,7 +820,10 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry,
 
 	if (victim_f) {
 		/* There was a victim. Kill it off nicely */
-		drop_nlink(new_dentry->d_inode);
+		if (S_ISDIR(new_dentry->d_inode->i_mode))
+			clear_nlink(new_dentry->d_inode);
+		else
+			drop_nlink(new_dentry->d_inode);
 		/* Don't oops if the victim was a dirent pointing to an
 		   inode which didn't exist. */
 		if (victim_f->inocache) {
-- 
cgit v1.2.3


From d769b3c2ab7184ddd42056595b627cc871caa90e Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Thu, 21 Jul 2011 22:22:25 +0200
Subject: isofs: Remove global fs lock

sbi->s_mutex isn't needed for isofs at all so we can just remove it. Generally,
since isofs is always mounted read-only, filesystem structure cannot change
under us.  So buffer_head contents stays constant after it's filled in. That
leaves us with possible changes of global data structures. Superblock changes
only during filesystem mount (even remount does not change it), inodes are only
filled in during reading from disk. So there are no changes of these structures
to bother about.

Arguments why sbi->s_mutex can be removed at each place:
isofs_readdir: Accesses sb, inode, filp, local variables => s_mutex not needed
isofs_lookup: Protected by directory's i_mutex. Accesses sb, inode, dentry,
  local variables => s_mutex not needed
rock_ridge_symlink_readpage: Protected by page lock. Accesses sb, inode,
  local variables => s_mutex not needed.

Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/isofs/dir.c   | 3 ---
 fs/isofs/inode.c | 1 -
 fs/isofs/isofs.h | 1 -
 fs/isofs/namei.c | 4 ----
 fs/isofs/rock.c  | 3 ---
 5 files changed, 12 deletions(-)

(limited to 'fs')

diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c
index 0542b6eedf80..f20437c068a0 100644
--- a/fs/isofs/dir.c
+++ b/fs/isofs/dir.c
@@ -254,19 +254,16 @@ static int isofs_readdir(struct file *filp,
 	char *tmpname;
 	struct iso_directory_record *tmpde;
 	struct inode *inode = filp->f_path.dentry->d_inode;
-	struct isofs_sb_info *sbi = ISOFS_SB(inode->i_sb);
 
 	tmpname = (char *)__get_free_page(GFP_KERNEL);
 	if (tmpname == NULL)
 		return -ENOMEM;
 
-	mutex_lock(&sbi->s_mutex);
 	tmpde = (struct iso_directory_record *) (tmpname+1024);
 
 	result = do_isofs_readdir(inode, filp, dirent, filldir, tmpname, tmpde);
 
 	free_page((unsigned long) tmpname);
-	mutex_unlock(&sbi->s_mutex);
 	return result;
 }
 
diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c
index b3cc8586984e..a5d03672d04e 100644
--- a/fs/isofs/inode.c
+++ b/fs/isofs/inode.c
@@ -863,7 +863,6 @@ root_found:
 	sbi->s_utf8 = opt.utf8;
 	sbi->s_nocompress = opt.nocompress;
 	sbi->s_overriderockperm = opt.overriderockperm;
-	mutex_init(&sbi->s_mutex);
 	/*
 	 * It would be incredibly stupid to allow people to mark every file
 	 * on the disk as suid, so we merely allow them to set the default
diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h
index 2882dc089f87..7d33de84f52a 100644
--- a/fs/isofs/isofs.h
+++ b/fs/isofs/isofs.h
@@ -55,7 +55,6 @@ struct isofs_sb_info {
 	gid_t s_gid;
 	uid_t s_uid;
 	struct nls_table *s_nls_iocharset; /* Native language support table */
-	struct mutex s_mutex; /* replaces BKL, please remove if possible */
 };
 
 #define ISOFS_INVALID_MODE ((mode_t) -1)
diff --git a/fs/isofs/namei.c b/fs/isofs/namei.c
index 68fa503d877e..1e2946f2a69e 100644
--- a/fs/isofs/namei.c
+++ b/fs/isofs/namei.c
@@ -168,7 +168,6 @@ struct dentry *isofs_lookup(struct inode *dir, struct dentry *dentry, struct nam
 	int found;
 	unsigned long uninitialized_var(block);
 	unsigned long uninitialized_var(offset);
-	struct isofs_sb_info *sbi = ISOFS_SB(dir->i_sb);
 	struct inode *inode;
 	struct page *page;
 
@@ -176,7 +175,6 @@ struct dentry *isofs_lookup(struct inode *dir, struct dentry *dentry, struct nam
 	if (!page)
 		return ERR_PTR(-ENOMEM);
 
-	mutex_lock(&sbi->s_mutex);
 	found = isofs_find_entry(dir, dentry,
 				&block, &offset,
 				page_address(page),
@@ -185,7 +183,5 @@ struct dentry *isofs_lookup(struct inode *dir, struct dentry *dentry, struct nam
 
 	inode = found ? isofs_iget(dir->i_sb, block, offset) : NULL;
 
-	mutex_unlock(&sbi->s_mutex);
-
 	return d_splice_alias(inode, dentry);
 }
diff --git a/fs/isofs/rock.c b/fs/isofs/rock.c
index f9cd04db6eab..1fbc7de88f50 100644
--- a/fs/isofs/rock.c
+++ b/fs/isofs/rock.c
@@ -678,7 +678,6 @@ static int rock_ridge_symlink_readpage(struct file *file, struct page *page)
 
 	init_rock_state(&rs, inode);
 	block = ei->i_iget5_block;
-	mutex_lock(&sbi->s_mutex);
 	bh = sb_bread(inode->i_sb, block);
 	if (!bh)
 		goto out_noread;
@@ -748,7 +747,6 @@ repeat:
 		goto fail;
 	brelse(bh);
 	*rpnt = '\0';
-	mutex_unlock(&sbi->s_mutex);
 	SetPageUptodate(page);
 	kunmap(page);
 	unlock_page(page);
@@ -765,7 +763,6 @@ out_bad_span:
 	printk("symlink spans iso9660 blocks\n");
 fail:
 	brelse(bh);
-	mutex_unlock(&sbi->s_mutex);
 error:
 	SetPageError(page);
 	kunmap(page);
-- 
cgit v1.2.3


From 5a9a43646cf709312d71eca71cef90ad802f28f9 Mon Sep 17 00:00:00 2001
From: Konstantin Khlebnikov <khlebnikov@openvz.org>
Date: Sun, 17 Jul 2011 15:35:12 +0400
Subject: vfs: use ERR_CAST for err-ptr tossing in lookup_instantiate_filp

Replace unclear (struct dentry *) to (struct file *) typecast with ERR_CAST() macro.

Signed-off-by: Konstantin Khlebnikov <khlebnikov@openvz.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/open.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/open.c b/fs/open.c
index b52cf013ffa1..739b751aa73e 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -793,7 +793,7 @@ out:
 	return nd->intent.open.file;
 out_err:
 	release_open_intent(nd);
-	nd->intent.open.file = (struct file *)dentry;
+	nd->intent.open.file = ERR_CAST(dentry);
 	goto out;
 }
 EXPORT_SYMBOL_GPL(lookup_instantiate_filp);
-- 
cgit v1.2.3