From ac45d61357e86b9a0cf14e45e8e09dfb626970ef Mon Sep 17 00:00:00 2001
From: Miklos Szeredi <mszeredi@suse.cz>
Date: Mon, 5 Mar 2012 15:48:11 +0100
Subject: fuse: fix nlink after unlink

Anand Avati reports that the following sequence of system calls fail on a fuse
filesystem:


 	create("filename") => 0
 	link("filename", "linkname") => 0
 	unlink("filename") => 0
 	link("linkname", "filename") => -ENOENT ### BUG ###

vfs_link() fails with ENOENT if i_nlink is zero, this is done to prevent
resurrecting already deleted files.

Fuse clears i_nlink on unlink even if there are other links pointing to the
file.

Reported-by: Anand Avati <avati@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
---
 fs/fuse/dir.c | 22 +++++++++++++++-------
 1 file changed, 15 insertions(+), 7 deletions(-)

(limited to 'fs')

diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 206632887bb..da379070fab 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -644,13 +644,12 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry)
 	fuse_put_request(fc, req);
 	if (!err) {
 		struct inode *inode = entry->d_inode;
+		struct fuse_inode *fi = get_fuse_inode(inode);
 
-		/*
-		 * Set nlink to zero so the inode can be cleared, if the inode
-		 * does have more links this will be discovered at the next
-		 * lookup/getattr.
-		 */
-		clear_nlink(inode);
+		spin_lock(&fc->lock);
+		fi->attr_version = ++fc->attr_version;
+		drop_nlink(inode);
+		spin_unlock(&fc->lock);
 		fuse_invalidate_attr(inode);
 		fuse_invalidate_attr(dir);
 		fuse_invalidate_entry_cache(entry);
@@ -762,8 +761,17 @@ static int fuse_link(struct dentry *entry, struct inode *newdir,
 	   will reflect changes in the backing inode (link count,
 	   etc.)
 	*/
-	if (!err || err == -EINTR)
+	if (!err) {
+		struct fuse_inode *fi = get_fuse_inode(inode);
+
+		spin_lock(&fc->lock);
+		fi->attr_version = ++fc->attr_version;
+		inc_nlink(inode);
+		spin_unlock(&fc->lock);
 		fuse_invalidate_attr(inode);
+	} else if (err == -EINTR) {
+		fuse_invalidate_attr(inode);
+	}
 	return err;
 }
 
-- 
cgit v1.2.3


From 4273b793ec68753cc3fcf5be7cbfd88c2be2058d Mon Sep 17 00:00:00 2001
From: Anand Avati <avati@redhat.com>
Date: Fri, 17 Feb 2012 12:46:25 -0500
Subject: fuse: O_DIRECT support for files

Implement ->direct_IO() method in aops. The ->direct_IO() method combines
the existing fuse_direct_read/fuse_direct_write methods to implement
O_DIRECT functionality.

Reaching ->direct_IO() in the read path via generic_file_aio_read ensures
proper synchronization with page cache with its existing framework.

Reaching ->direct_IO() in the write path via fuse_file_aio_write is made
to come via generic_file_direct_write() which makes it play nice with
the page cache w.r.t other mmap pages etc.

On files marked 'direct_io' by the filesystem server, IO always follows
the fuse_direct_read/write path. There is no effect of fcntl(O_DIRECT)
and it always succeeds.

On files not marked with 'direct_io' by the filesystem server, the IO
path depends on O_DIRECT flag by the application. This can be passed
at the time of open() as well as via fcntl().

Note that asynchronous O_DIRECT iocb jobs are completed synchronously
always (this has been the case with FUSE even before this patch)

Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
---
 fs/fuse/dir.c  |   3 --
 fs/fuse/file.c | 129 +++++++++++++++++++++++++++++++++++++++++++++++++--------
 2 files changed, 112 insertions(+), 20 deletions(-)

(limited to 'fs')

diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index da379070fab..df5ac048dc7 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -387,9 +387,6 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry,
 	if (fc->no_create)
 		return -ENOSYS;
 
-	if (flags & O_DIRECT)
-		return -EINVAL;
-
 	forget = fuse_alloc_forget();
 	if (!forget)
 		return -ENOMEM;
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index 4a199fd93fb..8ca007d09c9 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -194,10 +194,6 @@ int fuse_open_common(struct inode *inode, struct file *file, bool isdir)
 	struct fuse_conn *fc = get_fuse_conn(inode);
 	int err;
 
-	/* VFS checks this, but only _after_ ->open() */
-	if (file->f_flags & O_DIRECT)
-		return -EINVAL;
-
 	err = generic_file_open(inode, file);
 	if (err)
 		return err;
@@ -932,17 +928,23 @@ static ssize_t fuse_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
 	struct file *file = iocb->ki_filp;
 	struct address_space *mapping = file->f_mapping;
 	size_t count = 0;
+	size_t ocount = 0;
 	ssize_t written = 0;
+	ssize_t written_buffered = 0;
 	struct inode *inode = mapping->host;
 	ssize_t err;
 	struct iov_iter i;
+	loff_t endbyte = 0;
 
 	WARN_ON(iocb->ki_pos != pos);
 
-	err = generic_segment_checks(iov, &nr_segs, &count, VERIFY_READ);
+	ocount = 0;
+	err = generic_segment_checks(iov, &nr_segs, &ocount, VERIFY_READ);
 	if (err)
 		return err;
 
+	count = ocount;
+
 	mutex_lock(&inode->i_mutex);
 	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
 
@@ -962,11 +964,41 @@ static ssize_t fuse_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
 
 	file_update_time(file);
 
-	iov_iter_init(&i, iov, nr_segs, count, 0);
-	written = fuse_perform_write(file, mapping, &i, pos);
-	if (written >= 0)
-		iocb->ki_pos = pos + written;
+	if (file->f_flags & O_DIRECT) {
+		written = generic_file_direct_write(iocb, iov, &nr_segs,
+						    pos, &iocb->ki_pos,
+						    count, ocount);
+		if (written < 0 || written == count)
+			goto out;
+
+		pos += written;
+		count -= written;
 
+		iov_iter_init(&i, iov, nr_segs, count, written);
+		written_buffered = fuse_perform_write(file, mapping, &i, pos);
+		if (written_buffered < 0) {
+			err = written_buffered;
+			goto out;
+		}
+		endbyte = pos + written_buffered - 1;
+
+		err = filemap_write_and_wait_range(file->f_mapping, pos,
+						   endbyte);
+		if (err)
+			goto out;
+
+		invalidate_mapping_pages(file->f_mapping,
+					 pos >> PAGE_CACHE_SHIFT,
+					 endbyte >> PAGE_CACHE_SHIFT);
+
+		written += written_buffered;
+		iocb->ki_pos = pos + written_buffered;
+	} else {
+		iov_iter_init(&i, iov, nr_segs, count, 0);
+		written = fuse_perform_write(file, mapping, &i, pos);
+		if (written >= 0)
+			iocb->ki_pos = pos + written;
+	}
 out:
 	current->backing_dev_info = NULL;
 	mutex_unlock(&inode->i_mutex);
@@ -1101,30 +1133,41 @@ static ssize_t fuse_direct_read(struct file *file, char __user *buf,
 	return res;
 }
 
-static ssize_t fuse_direct_write(struct file *file, const char __user *buf,
-				 size_t count, loff_t *ppos)
+static ssize_t __fuse_direct_write(struct file *file, const char __user *buf,
+				   size_t count, loff_t *ppos)
 {
 	struct inode *inode = file->f_path.dentry->d_inode;
 	ssize_t res;
 
-	if (is_bad_inode(inode))
-		return -EIO;
-
-	/* Don't allow parallel writes to the same file */
-	mutex_lock(&inode->i_mutex);
 	res = generic_write_checks(file, ppos, &count, 0);
 	if (!res) {
 		res = fuse_direct_io(file, buf, count, ppos, 1);
 		if (res > 0)
 			fuse_write_update_size(inode, *ppos);
 	}
-	mutex_unlock(&inode->i_mutex);
 
 	fuse_invalidate_attr(inode);
 
 	return res;
 }
 
+static ssize_t fuse_direct_write(struct file *file, const char __user *buf,
+				 size_t count, loff_t *ppos)
+{
+	struct inode *inode = file->f_path.dentry->d_inode;
+	ssize_t res;
+
+	if (is_bad_inode(inode))
+		return -EIO;
+
+	/* Don't allow parallel writes to the same file */
+	mutex_lock(&inode->i_mutex);
+	res = __fuse_direct_write(file, buf, count, ppos);
+	mutex_unlock(&inode->i_mutex);
+
+	return res;
+}
+
 static void fuse_writepage_free(struct fuse_conn *fc, struct fuse_req *req)
 {
 	__free_page(req->pages[0]);
@@ -2077,6 +2120,57 @@ int fuse_notify_poll_wakeup(struct fuse_conn *fc,
 	return 0;
 }
 
+static ssize_t fuse_loop_dio(struct file *filp, const struct iovec *iov,
+			     unsigned long nr_segs, loff_t *ppos, int rw)
+{
+	const struct iovec *vector = iov;
+	ssize_t ret = 0;
+
+	while (nr_segs > 0) {
+		void __user *base;
+		size_t len;
+		ssize_t nr;
+
+		base = vector->iov_base;
+		len = vector->iov_len;
+		vector++;
+		nr_segs--;
+
+		if (rw == WRITE)
+			nr = __fuse_direct_write(filp, base, len, ppos);
+		else
+			nr = fuse_direct_read(filp, base, len, ppos);
+
+		if (nr < 0) {
+			if (!ret)
+				ret = nr;
+			break;
+		}
+		ret += nr;
+		if (nr != len)
+			break;
+	}
+
+	return ret;
+}
+
+
+static ssize_t
+fuse_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
+			loff_t offset, unsigned long nr_segs)
+{
+	ssize_t ret = 0;
+	struct file *file = NULL;
+	loff_t pos = 0;
+
+	file = iocb->ki_filp;
+	pos = offset;
+
+	ret = fuse_loop_dio(file, iov, nr_segs, &pos, rw);
+
+	return ret;
+}
+
 static const struct file_operations fuse_file_operations = {
 	.llseek		= fuse_file_llseek,
 	.read		= do_sync_read,
@@ -2120,6 +2214,7 @@ static const struct address_space_operations fuse_file_aops  = {
 	.readpages	= fuse_readpages,
 	.set_page_dirty	= __set_page_dirty_nobuffers,
 	.bmap		= fuse_bmap,
+	.direct_IO	= fuse_direct_IO,
 };
 
 void fuse_init_file_inode(struct inode *inode)
-- 
cgit v1.2.3


From 0a2da9b2ef2ef76c09397597f260245b020e6522 Mon Sep 17 00:00:00 2001
From: Miklos Szeredi <mszeredi@suse.cz>
Date: Wed, 11 Apr 2012 11:45:06 +0200
Subject: fuse: allow nanosecond granularity

Derrik Pates reports that an utimensat with a NULL argument results in the
current time being sent from the kernel with 1 second granularity.

Reported-by: Derrik Pates <demon@now.ai>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
---
 fs/fuse/inode.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'fs')

diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index 64cf8d07393..c9a0c972a4b 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -947,6 +947,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent)
 	sb->s_magic = FUSE_SUPER_MAGIC;
 	sb->s_op = &fuse_super_operations;
 	sb->s_maxbytes = MAX_LFS_FILESIZE;
+	sb->s_time_gran = 1;
 	sb->s_export_op = &fuse_export_operations;
 
 	file = fget(d.fd);
-- 
cgit v1.2.3


From 9dc4e6c4d1182d34604ea40fef641775f5b15456 Mon Sep 17 00:00:00 2001
From: "J. Bruce Fields" <bfields@redhat.com>
Date: Mon, 9 Apr 2012 18:06:49 -0400
Subject: nfsd: don't fail unchecked creates of non-special files

Allow a v3 unchecked open of a non-regular file succeed as if it were a
lookup; typically a client in such a case will want to fall back on a
local open, so succeeding and giving it the filehandle is more useful
than failing with nfserr_exist, which makes it appear that nothing at
all exists by that name.

Similarly for v4, on an open-create, return the same errors we would on
an attempt to open a non-regular file, instead of returning
nfserr_exist.

This fixes a problem found doing a v4 open of a symlink with
O_RDONLY|O_CREAT, which resulted in the current client returning EEXIST.

Thanks also to Trond for analysis.

Cc: stable@kernel.org
Reported-by: Orion Poplawski <orion@cora.nwra.com>
Tested-by: Orion Poplawski <orion@cora.nwra.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
---
 fs/nfsd/nfs4proc.c | 8 ++++----
 fs/nfsd/vfs.c      | 2 +-
 2 files changed, 5 insertions(+), 5 deletions(-)

(limited to 'fs')

diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 2ed14dfd00a..8256efda363 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -235,15 +235,15 @@ do_open_lookup(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfsd4_o
 		 */
 		if (open->op_createmode == NFS4_CREATE_EXCLUSIVE && status == 0)
 			open->op_bmval[1] = (FATTR4_WORD1_TIME_ACCESS |
-						FATTR4_WORD1_TIME_MODIFY);
+							FATTR4_WORD1_TIME_MODIFY);
 	} else {
 		status = nfsd_lookup(rqstp, current_fh,
 				     open->op_fname.data, open->op_fname.len, resfh);
 		fh_unlock(current_fh);
-		if (status)
-			goto out;
-		status = nfsd_check_obj_isreg(resfh);
 	}
+	if (status)
+		goto out;
+	status = nfsd_check_obj_isreg(resfh);
 	if (status)
 		goto out;
 
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 296d671654d..568666156ea 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1458,7 +1458,7 @@ do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
 		switch (createmode) {
 		case NFS3_CREATE_UNCHECKED:
 			if (! S_ISREG(dchild->d_inode->i_mode))
-				err = nfserr_exist;
+				goto out;
 			else if (truncp) {
 				/* in nfsv4, we need to treat this case a little
 				 * differently.  we don't want to truncate the
-- 
cgit v1.2.3


From 4fe9e9639d95cd11de63afa353f2de320f26033a Mon Sep 17 00:00:00 2001
From: Sachin Prabhu <sprabhu@redhat.com>
Date: Tue, 10 Apr 2012 18:12:27 +0100
Subject: Cleanup handling of NULL value passed for a mount option

Allow blank user= and ip= mount option. Also clean up redundant
checks for NULL values since the token parser will not actually
match mount options with NULL values unless explicitly specified.

Signed-off-by: Sachin Prabhu <sprabhu@redhat.com>
Reported-by: Chris Clayton <chris2553@googlemail.com>
Acked-by: Jeff Layton <jlayton@samba.org>
Tested-by: Chris Clayton <chris2553@googlemail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
---
 fs/cifs/connect.c | 80 +++++++++++++------------------------------------------
 1 file changed, 19 insertions(+), 61 deletions(-)

(limited to 'fs')

diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index d81e933a796..6a86f3d6818 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -109,6 +109,8 @@ enum {
 
 	/* Options which could be blank */
 	Opt_blank_pass,
+	Opt_blank_user,
+	Opt_blank_ip,
 
 	Opt_err
 };
@@ -183,11 +185,15 @@ static const match_table_t cifs_mount_option_tokens = {
 	{ Opt_wsize, "wsize=%s" },
 	{ Opt_actimeo, "actimeo=%s" },
 
+	{ Opt_blank_user, "user=" },
+	{ Opt_blank_user, "username=" },
 	{ Opt_user, "user=%s" },
 	{ Opt_user, "username=%s" },
 	{ Opt_blank_pass, "pass=" },
 	{ Opt_pass, "pass=%s" },
 	{ Opt_pass, "password=%s" },
+	{ Opt_blank_ip, "ip=" },
+	{ Opt_blank_ip, "addr=" },
 	{ Opt_ip, "ip=%s" },
 	{ Opt_ip, "addr=%s" },
 	{ Opt_unc, "unc=%s" },
@@ -1534,15 +1540,17 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 
 		/* String Arguments */
 
+		case Opt_blank_user:
+			/* null user, ie. anonymous authentication */
+			vol->nullauth = 1;
+			vol->username = NULL;
+			break;
 		case Opt_user:
 			string = match_strdup(args);
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				/* null user, ie. anonymous authentication */
-				vol->nullauth = 1;
-			} else if (strnlen(string, MAX_USERNAME_SIZE) >
+			if (strnlen(string, MAX_USERNAME_SIZE) >
 							MAX_USERNAME_SIZE) {
 				printk(KERN_WARNING "CIFS: username too long\n");
 				goto cifs_parse_mount_err;
@@ -1611,14 +1619,15 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			}
 			vol->password[j] = '\0';
 			break;
+		case Opt_blank_ip:
+			vol->UNCip = NULL;
+			break;
 		case Opt_ip:
 			string = match_strdup(args);
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				vol->UNCip = NULL;
-			} else if (strnlen(string, INET6_ADDRSTRLEN) >
+			if (strnlen(string, INET6_ADDRSTRLEN) >
 						INET6_ADDRSTRLEN) {
 				printk(KERN_WARNING "CIFS: ip address "
 						    "too long\n");
@@ -1636,12 +1645,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: invalid path to "
-						    "network resource\n");
-				goto cifs_parse_mount_err;
-			}
-
 			temp_len = strnlen(string, 300);
 			if (temp_len  == 300) {
 				printk(KERN_WARNING "CIFS: UNC name too long\n");
@@ -1670,11 +1673,7 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: invalid domain"
-						    " name\n");
-				goto cifs_parse_mount_err;
-			} else if (strnlen(string, 256) == 256) {
+			if (strnlen(string, 256) == 256) {
 				printk(KERN_WARNING "CIFS: domain name too"
 						    " long\n");
 				goto cifs_parse_mount_err;
@@ -1693,11 +1692,7 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: srcaddr value not"
-						    " specified\n");
-				goto cifs_parse_mount_err;
-			} else if (!cifs_convert_address(
+			if (!cifs_convert_address(
 					(struct sockaddr *)&vol->srcaddr,
 					string, strlen(string))) {
 				printk(KERN_WARNING "CIFS:  Could not parse"
@@ -1710,11 +1705,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: Invalid path"
-						    " prefix\n");
-				goto cifs_parse_mount_err;
-			}
 			temp_len = strnlen(string, 1024);
 			if (string[0] != '/')
 				temp_len++; /* missing leading slash */
@@ -1742,11 +1732,7 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: Invalid iocharset"
-						    " specified\n");
-				goto cifs_parse_mount_err;
-			} else if (strnlen(string, 1024) >= 65) {
+			if (strnlen(string, 1024) >= 65) {
 				printk(KERN_WARNING "CIFS: iocharset name "
 						    "too long.\n");
 				goto cifs_parse_mount_err;
@@ -1771,11 +1757,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: No socket option"
-						    " specified\n");
-				goto cifs_parse_mount_err;
-			}
 			if (strnicmp(string, "TCP_NODELAY", 11) == 0)
 				vol->sockopt_tcp_nodelay = 1;
 			break;
@@ -1784,12 +1765,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: Invalid (empty)"
-						    " netbiosname\n");
-				break;
-			}
-
 			memset(vol->source_rfc1001_name, 0x20,
 				RFC1001_NAME_LEN);
 			/*
@@ -1817,11 +1792,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: Empty server"
-					" netbiosname specified\n");
-				break;
-			}
 			/* last byte, type, is 0x20 for servr type */
 			memset(vol->target_rfc1001_name, 0x20,
 				RFC1001_NAME_LEN_WITH_NULL);
@@ -1848,12 +1818,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				cERROR(1, "no protocol version specified"
-					  " after vers= mount option");
-				goto cifs_parse_mount_err;
-			}
-
 			if (strnicmp(string, "cifs", 4) == 0 ||
 			    strnicmp(string, "1", 1) == 0) {
 				/* This is the default */
@@ -1868,12 +1832,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname,
 			if (string == NULL)
 				goto out_nomem;
 
-			if (!*string) {
-				printk(KERN_WARNING "CIFS: no security flavor"
-						    " specified\n");
-				break;
-			}
-
 			if (cifs_parse_security_flavors(string, vol) != 0)
 				goto cifs_parse_mount_err;
 			break;
-- 
cgit v1.2.3


From 96f6f98501196d46ce52c2697dd758d9300c63f5 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 12 Apr 2012 23:47:00 -0400
Subject: nfsd: fix b0rken error value for setattr on read-only mount

..._want_write() returns -EROFS on failure, _not_ an NFS error value.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfsd/nfs4proc.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 2ed14dfd00a..694d526c8f1 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -841,6 +841,7 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	      struct nfsd4_setattr *setattr)
 {
 	__be32 status = nfs_ok;
+	int err;
 
 	if (setattr->sa_iattr.ia_valid & ATTR_SIZE) {
 		nfs4_lock_state();
@@ -852,9 +853,9 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 			return status;
 		}
 	}
-	status = fh_want_write(&cstate->current_fh);
-	if (status)
-		return status;
+	err = fh_want_write(&cstate->current_fh);
+	if (err)
+		return nfserrno(err);
 	status = nfs_ok;
 
 	status = check_attr_support(rqstp, cstate, setattr->sa_bmval,
-- 
cgit v1.2.3


From 04da6e9d63427b2d0fd04766712200c250b3278f Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 00:00:04 -0400
Subject: nfsd: fix error values returned by nfsd4_lockt() when nfsd_open()
 fails

nfsd_open() already returns an NFS error value; only vfs_test_lock()
result needs to be fed through nfserrno().  Broken by commit 55ef12
(nfsd: Ensure nfsv4 calls the underlying filesystem on LOCKT)
three years ago...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfsd/nfs4state.c | 23 +++++++++--------------
 1 file changed, 9 insertions(+), 14 deletions(-)

(limited to 'fs')

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 1841f8bf845..7f71c69cdcd 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -4211,16 +4211,14 @@ out:
  * vfs_test_lock.  (Arguably perhaps test_lock should be done with an
  * inode operation.)
  */
-static int nfsd_test_lock(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file_lock *lock)
+static __be32 nfsd_test_lock(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file_lock *lock)
 {
 	struct file *file;
-	int err;
-
-	err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file);
-	if (err)
-		return err;
-	err = vfs_test_lock(file, lock);
-	nfsd_close(file);
+	__be32 err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file);
+	if (!err) {
+		err = nfserrno(vfs_test_lock(file, lock));
+		nfsd_close(file);
+	}
 	return err;
 }
 
@@ -4234,7 +4232,6 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 	struct inode *inode;
 	struct file_lock file_lock;
 	struct nfs4_lockowner *lo;
-	int error;
 	__be32 status;
 
 	if (locks_in_grace())
@@ -4280,12 +4277,10 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
 
 	nfs4_transform_lock_offset(&file_lock);
 
-	status = nfs_ok;
-	error = nfsd_test_lock(rqstp, &cstate->current_fh, &file_lock);
-	if (error) {
-		status = nfserrno(error);
+	status = nfsd_test_lock(rqstp, &cstate->current_fh, &file_lock);
+	if (status)
 		goto out;
-	}
+
 	if (file_lock.fl_type != F_UNLCK) {
 		status = nfserr_denied;
 		nfs4_set_lock_denied(&file_lock, &lockt->lt_denied);
-- 
cgit v1.2.3


From 02f5fde5df0ea930e70f93763dd48beff182b208 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 00:10:34 -0400
Subject: nfsd: fix endianness breakage in TEST_STATEID handling

->ts_id_status gets nfs errno, i.e. it's already big-endian; no need
to apply htonl() to it.  Broken by commit 174568 (NFSD: Added TEST_STATEID
operation) last year...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfsd/nfs4xdr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index bcd8904ab1e..07a99b6d4bb 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -3410,7 +3410,7 @@ nfsd4_encode_test_stateid(struct nfsd4_compoundres *resp, int nfserr,
 	*p++ = htonl(test_stateid->ts_num_ids);
 
 	list_for_each_entry_safe(stateid, next, &test_stateid->ts_stateid_list, ts_id_list) {
-		*p++ = htonl(stateid->ts_id_status);
+		*p++ = stateid->ts_id_status;
 	}
 
 	ADJUST_ARGS();
-- 
cgit v1.2.3


From afcf6792afd66209161495f691e19d4fc5460a93 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 00:15:37 -0400
Subject: nfsd: fix error value on allocation failure in
 nfsd4_decode_test_stateid()

PTR_ERR(NULL) is going to be 0...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfsd/nfs4xdr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index 07a99b6d4bb..74c00bc92b9 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -1392,7 +1392,7 @@ nfsd4_decode_test_stateid(struct nfsd4_compoundargs *argp, struct nfsd4_test_sta
 	for (i = 0; i < test_stateid->ts_num_ids; i++) {
 		stateid = kmalloc(sizeof(struct nfsd4_test_stateid_id), GFP_KERNEL);
 		if (!stateid) {
-			status = PTR_ERR(stateid);
+			status = nfserrno(-ENOMEM);
 			goto out;
 		}
 
-- 
cgit v1.2.3


From efe39651f08813180f37dc508d950fc7d92b29a8 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 00:32:14 -0400
Subject: nfsd: fix compose_entry_fh() failure exits

Restore the original logics ("fail on mountpoints, negatives and in
case of fh_compose() failures").  Since commit 8177e (nfsd: clean up
readdirplus encoding) that got broken -
	rv = fh_compose(fhp, exp, dchild, &cd->fh);
	if (rv)
	       goto out;
	if (!dchild->d_inode)
		goto out;
	rv = 0;
out:
is equivalent to
	rv = fh_compose(fhp, exp, dchild, &cd->fh);
out:
and the second check has no effect whatsoever...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/nfsd/nfs3xdr.c | 22 ++++++++--------------
 1 file changed, 8 insertions(+), 14 deletions(-)

(limited to 'fs')

diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c
index 08c6e36ab2e..43f46cd9ede 100644
--- a/fs/nfsd/nfs3xdr.c
+++ b/fs/nfsd/nfs3xdr.c
@@ -803,13 +803,13 @@ encode_entry_baggage(struct nfsd3_readdirres *cd, __be32 *p, const char *name,
 	return p;
 }
 
-static int
+static __be32
 compose_entry_fh(struct nfsd3_readdirres *cd, struct svc_fh *fhp,
 		const char *name, int namlen)
 {
 	struct svc_export	*exp;
 	struct dentry		*dparent, *dchild;
-	int rv = 0;
+	__be32 rv = nfserr_noent;
 
 	dparent = cd->fh.fh_dentry;
 	exp  = cd->fh.fh_export;
@@ -817,26 +817,20 @@ compose_entry_fh(struct nfsd3_readdirres *cd, struct svc_fh *fhp,
 	if (isdotent(name, namlen)) {
 		if (namlen == 2) {
 			dchild = dget_parent(dparent);
-			if (dchild == dparent) {
-				/* filesystem root - cannot return filehandle for ".." */
-				dput(dchild);
-				return -ENOENT;
-			}
+			/* filesystem root - cannot return filehandle for ".." */
+			if (dchild == dparent)
+				goto out;
 		} else
 			dchild = dget(dparent);
 	} else
 		dchild = lookup_one_len(name, dparent, namlen);
 	if (IS_ERR(dchild))
-		return -ENOENT;
-	rv = -ENOENT;
+		return rv;
 	if (d_mountpoint(dchild))
 		goto out;
-	rv = fh_compose(fhp, exp, dchild, &cd->fh);
-	if (rv)
-		goto out;
 	if (!dchild->d_inode)
 		goto out;
-	rv = 0;
+	rv = fh_compose(fhp, exp, dchild, &cd->fh);
 out:
 	dput(dchild);
 	return rv;
@@ -845,7 +839,7 @@ out:
 static __be32 *encode_entryplus_baggage(struct nfsd3_readdirres *cd, __be32 *p, const char *name, int namlen)
 {
 	struct svc_fh	fh;
-	int err;
+	__be32 err;
 
 	fh_init(&fh, NFS3_FHSIZE);
 	err = compose_entry_fh(cd, &fh, name, namlen);
-- 
cgit v1.2.3


From af1584f570b19b0285e4402a0b54731495d31784 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Thu, 12 Apr 2012 20:32:25 -0400
Subject: ext4: fix endianness breakage in ext4_split_extent_at()

->ee_len is __le16, so assigning cpu_to_le32() to it is going to do
Bad Things(tm) on big-endian hosts...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ext4/extents.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 1421938e679..1d418387ffb 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2882,7 +2882,7 @@ static int ext4_split_extent_at(handle_t *handle,
 		if (err)
 			goto fix_extent_len;
 		/* update the extent length and mark as initialized */
-		ex->ee_len = cpu_to_le32(ee_len);
+		ex->ee_len = cpu_to_le16(ee_len);
 		ext4_ext_try_to_merge(inode, path, ex);
 		err = ext4_ext_dirty(handle, inode, path + depth);
 		goto out;
-- 
cgit v1.2.3


From bfa890a3cdeed29eef53d54cd7f80cec0fd46b11 Mon Sep 17 00:00:00 2001
From: Sachin Prabhu <sprabhu@redhat.com>
Date: Fri, 13 Apr 2012 14:04:32 +0100
Subject: Fix number parsing in cifs_parse_mount_options

The function kstrtoul() used to parse number strings in the mount
option parser is set to expect a base 10 number . This treats the octal
numbers passed for mount options such as file_mode as base10 numbers
leading to incorrect behavior.

Change the 'base' argument passed to kstrtoul from 10 to 0 to
allow it to auto-detect the base of the number passed.

Signed-off-by: Sachin Prabhu <sprabhu@redhat.com>
Acked-by: Jeff Layton <jlayton@samba.org>
Reported-by: Chris Clayton <chris2553@googlemail.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
---
 fs/cifs/connect.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index 6a86f3d6818..f31dc9ac37b 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -1123,7 +1123,7 @@ static int get_option_ul(substring_t args[], unsigned long *option)
 	string = match_strdup(args);
 	if (string == NULL)
 		return -ENOMEM;
-	rc = kstrtoul(string, 10, option);
+	rc = kstrtoul(string, 0, option);
 	kfree(string);
 
 	return rc;
-- 
cgit v1.2.3


From 6ed3cf2cdfce4c9f1d73171bd3f27d9cb77b734e Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 11:49:04 -0400
Subject: btrfs: btrfs_root_readonly() broken on big-endian

->root_flags is __le64 and all accesses to it go through the helpers
that do proper conversions.  Except for btrfs_root_readonly(), which
checks bit 0 as in host-endian...

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

(limited to 'fs')

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 5b8ef8eb352..3f65a812e28 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2166,7 +2166,7 @@ BTRFS_SETGET_STACK_FUNCS(root_last_snapshot, struct btrfs_root_item,
 
 static inline bool btrfs_root_readonly(struct btrfs_root *root)
 {
-	return root->root_item.flags & BTRFS_ROOT_SUBVOL_RDONLY;
+	return (root->root_item.flags & cpu_to_le64(BTRFS_ROOT_SUBVOL_RDONLY)) != 0;
 }
 
 /* struct btrfs_root_backup */
-- 
cgit v1.2.3


From 3a251f04fe97c3d335b745c98e4b377e3c3899f2 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 12:22:00 -0400
Subject: ocfs2: ->l_next_free_req breakage on big-endian

It's le16, not le32...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ocfs2/alloc.c        | 2 +-
 fs/ocfs2/refcounttree.c | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'fs')

diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index 3165aebb43c..31b9463fba1 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -1134,7 +1134,7 @@ static int ocfs2_adjust_rightmost_branch(handle_t *handle,
 	}
 
 	el = path_leaf_el(path);
-	rec = &el->l_recs[le32_to_cpu(el->l_next_free_rec) - 1];
+	rec = &el->l_recs[le16_to_cpu(el->l_next_free_rec) - 1];
 
 	ocfs2_adjust_rightmost_records(handle, et, path, rec);
 
diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index cf782338266..a7b7217062b 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -1036,14 +1036,14 @@ static int ocfs2_get_refcount_cpos_end(struct ocfs2_caching_info *ci,
 
 	tmp_el = left_path->p_node[subtree_root].el;
 	blkno = left_path->p_node[subtree_root+1].bh->b_blocknr;
-	for (i = 0; i < le32_to_cpu(tmp_el->l_next_free_rec); i++) {
+	for (i = 0; i < le16_to_cpu(tmp_el->l_next_free_rec); i++) {
 		if (le64_to_cpu(tmp_el->l_recs[i].e_blkno) == blkno) {
 			*cpos_end = le32_to_cpu(tmp_el->l_recs[i+1].e_cpos);
 			break;
 		}
 	}
 
-	BUG_ON(i == le32_to_cpu(tmp_el->l_next_free_rec));
+	BUG_ON(i == le16_to_cpu(tmp_el->l_next_free_rec));
 
 out:
 	ocfs2_free_path(left_path);
-- 
cgit v1.2.3


From e1bf4cc620fd143766ddfcee3b004a1d1bb34fd0 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 12:27:11 -0400
Subject: ocfs: ->rl_used breakage on big-endian

it's le16, not le32 or le64...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ocfs2/refcounttree.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

(limited to 'fs')

diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index a7b7217062b..bc90ebcec16 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -1468,7 +1468,7 @@ static int ocfs2_divide_leaf_refcount_block(struct buffer_head *ref_leaf_bh,
 
 	trace_ocfs2_divide_leaf_refcount_block(
 		(unsigned long long)ref_leaf_bh->b_blocknr,
-		le32_to_cpu(rl->rl_count), le32_to_cpu(rl->rl_used));
+		le32_to_cpu(rl->rl_count), le16_to_cpu(rl->rl_used));
 
 	/*
 	 * XXX: Improvement later.
@@ -2411,7 +2411,7 @@ static int ocfs2_calc_refcount_meta_credits(struct super_block *sb,
 				rb = (struct ocfs2_refcount_block *)
 							prev_bh->b_data;
 
-				if (le64_to_cpu(rb->rf_records.rl_used) +
+				if (le16_to_cpu(rb->rf_records.rl_used) +
 				    recs_add >
 				    le16_to_cpu(rb->rf_records.rl_count))
 					ref_blocks++;
@@ -2476,7 +2476,7 @@ static int ocfs2_calc_refcount_meta_credits(struct super_block *sb,
 	if (prev_bh) {
 		rb = (struct ocfs2_refcount_block *)prev_bh->b_data;
 
-		if (le64_to_cpu(rb->rf_records.rl_used) + recs_add >
+		if (le16_to_cpu(rb->rf_records.rl_used) + recs_add >
 		    le16_to_cpu(rb->rf_records.rl_count))
 			ref_blocks++;
 
@@ -3629,7 +3629,7 @@ int ocfs2_refcounted_xattr_delete_need(struct inode *inode,
 			 * one will split a refcount rec, so totally we need
 			 * clusters * 2 new refcount rec.
 			 */
-			if (le64_to_cpu(rb->rf_records.rl_used) + clusters * 2 >
+			if (le16_to_cpu(rb->rf_records.rl_used) + clusters * 2 >
 			    le16_to_cpu(rb->rf_records.rl_count))
 				ref_blocks++;
 
-- 
cgit v1.2.3


From 28748b325dc2d730ccc312830a91c4ae0c0d9379 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 12:28:21 -0400
Subject: ocfs2: ->rl_count endianness breakage

le16, not le32...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ocfs2/refcounttree.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index bc90ebcec16..9f32d7cbb7a 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -1468,7 +1468,7 @@ static int ocfs2_divide_leaf_refcount_block(struct buffer_head *ref_leaf_bh,
 
 	trace_ocfs2_divide_leaf_refcount_block(
 		(unsigned long long)ref_leaf_bh->b_blocknr,
-		le32_to_cpu(rl->rl_count), le16_to_cpu(rl->rl_used));
+		le16_to_cpu(rl->rl_count), le16_to_cpu(rl->rl_used));
 
 	/*
 	 * XXX: Improvement later.
-- 
cgit v1.2.3


From 72094e43e3af5020510f920321d71f1798fa896d Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 12:30:02 -0400
Subject: ocfs2: ->e_leaf_clusters endianness breakage

le16, not le32...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/ocfs2/suballoc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c
index ba5d97e4a73..f169da4624f 100644
--- a/fs/ocfs2/suballoc.c
+++ b/fs/ocfs2/suballoc.c
@@ -600,7 +600,7 @@ static void ocfs2_bg_alloc_cleanup(handle_t *handle,
 		ret = ocfs2_free_clusters(handle, cluster_ac->ac_inode,
 					  cluster_ac->ac_bh,
 					  le64_to_cpu(rec->e_blkno),
-					  le32_to_cpu(rec->e_leaf_clusters));
+					  le16_to_cpu(rec->e_leaf_clusters));
 		if (ret)
 			mlog_errno(ret);
 		/* Try all the clusters to free */
@@ -1628,7 +1628,7 @@ static int ocfs2_bg_discontig_fix_by_rec(struct ocfs2_suballoc_result *res,
 {
 	unsigned int bpc = le16_to_cpu(cl->cl_bpc);
 	unsigned int bitoff = le32_to_cpu(rec->e_cpos) * bpc;
-	unsigned int bitcount = le32_to_cpu(rec->e_leaf_clusters) * bpc;
+	unsigned int bitcount = le16_to_cpu(rec->e_leaf_clusters) * bpc;
 
 	if (res->sr_bit_offset < bitoff)
 		return 0;
-- 
cgit v1.2.3


From e847469bf77a1d339274074ed068d461f0c872bc Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 13 Apr 2012 13:49:47 -0400
Subject: lockd: fix the endianness bug

comparing be32 values for < is not doing the right thing...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/lockd/clnt4xdr.c | 2 +-
 fs/lockd/clntxdr.c  | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'fs')

diff --git a/fs/lockd/clnt4xdr.c b/fs/lockd/clnt4xdr.c
index 3ddcbb1c0a4..13ad1539fbf 100644
--- a/fs/lockd/clnt4xdr.c
+++ b/fs/lockd/clnt4xdr.c
@@ -241,7 +241,7 @@ static int decode_nlm4_stat(struct xdr_stream *xdr, __be32 *stat)
 	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		goto out_overflow;
-	if (unlikely(*p > nlm4_failed))
+	if (unlikely(ntohl(*p) > ntohl(nlm4_failed)))
 		goto out_bad_xdr;
 	*stat = *p;
 	return 0;
diff --git a/fs/lockd/clntxdr.c b/fs/lockd/clntxdr.c
index 3d35e3e80c1..d269ada7670 100644
--- a/fs/lockd/clntxdr.c
+++ b/fs/lockd/clntxdr.c
@@ -236,7 +236,7 @@ static int decode_nlm_stat(struct xdr_stream *xdr,
 	p = xdr_inline_decode(xdr, 4);
 	if (unlikely(p == NULL))
 		goto out_overflow;
-	if (unlikely(*p > nlm_lck_denied_grace_period))
+	if (unlikely(ntohl(*p) > ntohl(nlm_lck_denied_grace_period)))
 		goto out_enum;
 	*stat = *p;
 	return 0;
-- 
cgit v1.2.3


From 9cd70b347e9761ea2d2ac3d758c529a48a8193e6 Mon Sep 17 00:00:00 2001
From: Theodore Ts'o <tytso@mit.edu>
Date: Mon, 16 Apr 2012 12:16:20 -0400
Subject: ext4: address scalability issue by removing extent cache statistics

Andi Kleen and Tim Chen have reported that under certain circumstances
the extent cache statistics are causing scalability problems due to
cache line bounces.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@vger.kernel.org
---
 fs/ext4/ext4.h    |  3 ---
 fs/ext4/extents.c |  4 ----
 fs/ext4/super.c   | 16 ----------------
 3 files changed, 23 deletions(-)

(limited to 'fs')

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index ab2594a30f8..0e01e90add8 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -1203,9 +1203,6 @@ struct ext4_sb_info {
 	unsigned long s_ext_blocks;
 	unsigned long s_ext_extents;
 #endif
-	/* ext4 extent cache stats */
-	unsigned long extent_cache_hits;
-	unsigned long extent_cache_misses;
 
 	/* for buddy allocator */
 	struct ext4_group_info ***s_group_info;
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 1421938e679..8a12cd20767 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2066,10 +2066,6 @@ static int ext4_ext_check_cache(struct inode *inode, ext4_lblk_t block,
 		ret = 1;
 	}
 errout:
-	if (!ret)
-		sbi->extent_cache_misses++;
-	else
-		sbi->extent_cache_hits++;
 	trace_ext4_ext_in_cache(inode, block, ret);
 	spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
 	return ret;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index ceebaf853be..ef7db504426 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -2366,18 +2366,6 @@ static ssize_t lifetime_write_kbytes_show(struct ext4_attr *a,
 			  EXT4_SB(sb)->s_sectors_written_start) >> 1)));
 }
 
-static ssize_t extent_cache_hits_show(struct ext4_attr *a,
-				      struct ext4_sb_info *sbi, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%lu\n", sbi->extent_cache_hits);
-}
-
-static ssize_t extent_cache_misses_show(struct ext4_attr *a,
-					struct ext4_sb_info *sbi, char *buf)
-{
-	return snprintf(buf, PAGE_SIZE, "%lu\n", sbi->extent_cache_misses);
-}
-
 static ssize_t inode_readahead_blks_store(struct ext4_attr *a,
 					  struct ext4_sb_info *sbi,
 					  const char *buf, size_t count)
@@ -2435,8 +2423,6 @@ static struct ext4_attr ext4_attr_##name = __ATTR(name, mode, show, store)
 EXT4_RO_ATTR(delayed_allocation_blocks);
 EXT4_RO_ATTR(session_write_kbytes);
 EXT4_RO_ATTR(lifetime_write_kbytes);
-EXT4_RO_ATTR(extent_cache_hits);
-EXT4_RO_ATTR(extent_cache_misses);
 EXT4_ATTR_OFFSET(inode_readahead_blks, 0644, sbi_ui_show,
 		 inode_readahead_blks_store, s_inode_readahead_blks);
 EXT4_RW_ATTR_SBI_UI(inode_goal, s_inode_goal);
@@ -2452,8 +2438,6 @@ static struct attribute *ext4_attrs[] = {
 	ATTR_LIST(delayed_allocation_blocks),
 	ATTR_LIST(session_write_kbytes),
 	ATTR_LIST(lifetime_write_kbytes),
-	ATTR_LIST(extent_cache_hits),
-	ATTR_LIST(extent_cache_misses),
 	ATTR_LIST(inode_readahead_blks),
 	ATTR_LIST(inode_goal),
 	ATTR_LIST(mb_stats),
-- 
cgit v1.2.3


From 57f73c2c89a5d3b2ed87201c8100d1fa989a1a65 Mon Sep 17 00:00:00 2001
From: Theodore Ts'o <tytso@mit.edu>
Date: Mon, 16 Apr 2012 18:55:26 -0400
Subject: ext4: fix handling of journalled quota options

Commit 26092bf5 broke handling of journalled quota mount options by
trying to parse argument of every mount option as a number.  Fix this
by dealing with the quota options before we call match_int().

Thanks to Jan Kara for discovering this regression.

Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/ext4/super.c | 32 +++++++++++++++-----------------
 1 file changed, 15 insertions(+), 17 deletions(-)

(limited to 'fs')

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index ef7db504426..6da193564e4 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1305,20 +1305,20 @@ static int set_qf_name(struct super_block *sb, int qtype, substring_t *args)
 		ext4_msg(sb, KERN_ERR,
 			"Cannot change journaled "
 			"quota options when quota turned on");
-		return 0;
+		return -1;
 	}
 	qname = match_strdup(args);
 	if (!qname) {
 		ext4_msg(sb, KERN_ERR,
 			"Not enough memory for storing quotafile name");
-		return 0;
+		return -1;
 	}
 	if (sbi->s_qf_names[qtype] &&
 		strcmp(sbi->s_qf_names[qtype], qname)) {
 		ext4_msg(sb, KERN_ERR,
 			"%s quota file already specified", QTYPE2NAME(qtype));
 		kfree(qname);
-		return 0;
+		return -1;
 	}
 	sbi->s_qf_names[qtype] = qname;
 	if (strchr(sbi->s_qf_names[qtype], '/')) {
@@ -1326,7 +1326,7 @@ static int set_qf_name(struct super_block *sb, int qtype, substring_t *args)
 			"quotafile must be on filesystem root");
 		kfree(sbi->s_qf_names[qtype]);
 		sbi->s_qf_names[qtype] = NULL;
-		return 0;
+		return -1;
 	}
 	set_opt(sb, QUOTA);
 	return 1;
@@ -1341,7 +1341,7 @@ static int clear_qf_name(struct super_block *sb, int qtype)
 		sbi->s_qf_names[qtype]) {
 		ext4_msg(sb, KERN_ERR, "Cannot change journaled quota options"
 			" when quota turned on");
-		return 0;
+		return -1;
 	}
 	/*
 	 * The space will be released later when all options are confirmed
@@ -1450,6 +1450,16 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
 	const struct mount_opts *m;
 	int arg = 0;
 
+#ifdef CONFIG_QUOTA
+	if (token == Opt_usrjquota)
+		return set_qf_name(sb, USRQUOTA, &args[0]);
+	else if (token == Opt_grpjquota)
+		return set_qf_name(sb, GRPQUOTA, &args[0]);
+	else if (token == Opt_offusrjquota)
+		return clear_qf_name(sb, USRQUOTA);
+	else if (token == Opt_offgrpjquota)
+		return clear_qf_name(sb, GRPQUOTA);
+#endif
 	if (args->from && match_int(args, &arg))
 		return -1;
 	switch (token) {
@@ -1549,18 +1559,6 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token,
 				sbi->s_mount_opt |= m->mount_opt;
 			}
 #ifdef CONFIG_QUOTA
-		} else if (token == Opt_usrjquota) {
-			if (!set_qf_name(sb, USRQUOTA, &args[0]))
-				return -1;
-		} else if (token == Opt_grpjquota) {
-			if (!set_qf_name(sb, GRPQUOTA, &args[0]))
-				return -1;
-		} else if (token == Opt_offusrjquota) {
-			if (!clear_qf_name(sb, USRQUOTA))
-				return -1;
-		} else if (token == Opt_offgrpjquota) {
-			if (!clear_qf_name(sb, GRPQUOTA))
-				return -1;
 		} else if (m->flags & MOPT_QFMT) {
 			if (sb_any_quota_loaded(sb) &&
 			    sbi->s_jquota_fmt != m->mount_opt) {
-- 
cgit v1.2.3


From e4eb1ff61b323d6141614e5458a1f53c7046ff8e Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Fri, 20 Apr 2012 15:35:40 -0700
Subject: VM: add "vm_brk()" helper function

It does the same thing as "do_brk()", except it handles the VM locking
too.

It turns out that all external callers want that anyway, so we can make
do_brk() static to just mm/mmap.c while at it.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/x86/ia32/ia32_aout.c | 20 +++++---------------
 fs/binfmt_aout.c          | 20 +++++---------------
 fs/binfmt_elf.c           | 15 ++++-----------
 include/linux/mm.h        |  3 ++-
 mm/mmap.c                 | 16 ++++++++++++++--
 mm/nommu.c                |  2 +-
 6 files changed, 31 insertions(+), 45 deletions(-)

(limited to 'fs')

diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c
index d511d951a05..b6817ee9033 100644
--- a/arch/x86/ia32/ia32_aout.c
+++ b/arch/x86/ia32/ia32_aout.c
@@ -119,9 +119,7 @@ static void set_brk(unsigned long start, unsigned long end)
 	end = PAGE_ALIGN(end);
 	if (end <= start)
 		return;
-	down_write(&current->mm->mmap_sem);
-	do_brk(start, end - start);
-	up_write(&current->mm->mmap_sem);
+	vm_brk(start, end - start);
 }
 
 #ifdef CORE_DUMP
@@ -332,9 +330,7 @@ static int load_aout_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 		pos = 32;
 		map_size = ex.a_text+ex.a_data;
 
-		down_write(&current->mm->mmap_sem);
-		error = do_brk(text_addr & PAGE_MASK, map_size);
-		up_write(&current->mm->mmap_sem);
+		error = vm_brk(text_addr & PAGE_MASK, map_size);
 
 		if (error != (text_addr & PAGE_MASK)) {
 			send_sig(SIGKILL, current, 0);
@@ -373,9 +369,7 @@ static int load_aout_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 		if (!bprm->file->f_op->mmap || (fd_offset & ~PAGE_MASK) != 0) {
 			loff_t pos = fd_offset;
 
-			down_write(&current->mm->mmap_sem);
-			do_brk(N_TXTADDR(ex), ex.a_text+ex.a_data);
-			up_write(&current->mm->mmap_sem);
+			vm_brk(N_TXTADDR(ex), ex.a_text+ex.a_data);
 			bprm->file->f_op->read(bprm->file,
 					(char __user *)N_TXTADDR(ex),
 					ex.a_text+ex.a_data, &pos);
@@ -476,9 +470,7 @@ static int load_aout_library(struct file *file)
 			error_time = jiffies;
 		}
 #endif
-		down_write(&current->mm->mmap_sem);
-		do_brk(start_addr, ex.a_text + ex.a_data + ex.a_bss);
-		up_write(&current->mm->mmap_sem);
+		vm_brk(start_addr, ex.a_text + ex.a_data + ex.a_bss);
 
 		file->f_op->read(file, (char __user *)start_addr,
 			ex.a_text + ex.a_data, &pos);
@@ -503,9 +495,7 @@ static int load_aout_library(struct file *file)
 	len = PAGE_ALIGN(ex.a_text + ex.a_data);
 	bss = ex.a_text + ex.a_data + ex.a_bss;
 	if (bss > len) {
-		down_write(&current->mm->mmap_sem);
-		error = do_brk(start_addr + len, bss - len);
-		up_write(&current->mm->mmap_sem);
+		error = vm_brk(start_addr + len, bss - len);
 		retval = error;
 		if (error != start_addr + len)
 			goto out;
diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c
index 2eb12f13593..88527492b91 100644
--- a/fs/binfmt_aout.c
+++ b/fs/binfmt_aout.c
@@ -50,9 +50,7 @@ static int set_brk(unsigned long start, unsigned long end)
 	end = PAGE_ALIGN(end);
 	if (end > start) {
 		unsigned long addr;
-		down_write(&current->mm->mmap_sem);
-		addr = do_brk(start, end - start);
-		up_write(&current->mm->mmap_sem);
+		addr = vm_brk(start, end - start);
 		if (BAD_ADDR(addr))
 			return addr;
 	}
@@ -280,9 +278,7 @@ static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs)
 		pos = 32;
 		map_size = ex.a_text+ex.a_data;
 #endif
-		down_write(&current->mm->mmap_sem);
-		error = do_brk(text_addr & PAGE_MASK, map_size);
-		up_write(&current->mm->mmap_sem);
+		error = vm_brk(text_addr & PAGE_MASK, map_size);
 		if (error != (text_addr & PAGE_MASK)) {
 			send_sig(SIGKILL, current, 0);
 			return error;
@@ -313,9 +309,7 @@ static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs)
 
 		if (!bprm->file->f_op->mmap||((fd_offset & ~PAGE_MASK) != 0)) {
 			loff_t pos = fd_offset;
-			down_write(&current->mm->mmap_sem);
-			do_brk(N_TXTADDR(ex), ex.a_text+ex.a_data);
-			up_write(&current->mm->mmap_sem);
+			vm_brk(N_TXTADDR(ex), ex.a_text+ex.a_data);
 			bprm->file->f_op->read(bprm->file,
 					(char __user *)N_TXTADDR(ex),
 					ex.a_text+ex.a_data, &pos);
@@ -412,9 +406,7 @@ static int load_aout_library(struct file *file)
 			       "N_TXTOFF is not page aligned. Please convert library: %s\n",
 			       file->f_path.dentry->d_name.name);
 		}
-		down_write(&current->mm->mmap_sem);
-		do_brk(start_addr, ex.a_text + ex.a_data + ex.a_bss);
-		up_write(&current->mm->mmap_sem);
+		vm_brk(start_addr, ex.a_text + ex.a_data + ex.a_bss);
 		
 		file->f_op->read(file, (char __user *)start_addr,
 			ex.a_text + ex.a_data, &pos);
@@ -438,9 +430,7 @@ static int load_aout_library(struct file *file)
 	len = PAGE_ALIGN(ex.a_text + ex.a_data);
 	bss = ex.a_text + ex.a_data + ex.a_bss;
 	if (bss > len) {
-		down_write(&current->mm->mmap_sem);
-		error = do_brk(start_addr + len, bss - len);
-		up_write(&current->mm->mmap_sem);
+		error = vm_brk(start_addr + len, bss - len);
 		retval = error;
 		if (error != start_addr + len)
 			goto out;
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 48ffb3dc610..0708a0bf0ba 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -82,9 +82,7 @@ static int set_brk(unsigned long start, unsigned long end)
 	end = ELF_PAGEALIGN(end);
 	if (end > start) {
 		unsigned long addr;
-		down_write(&current->mm->mmap_sem);
-		addr = do_brk(start, end - start);
-		up_write(&current->mm->mmap_sem);
+		addr = vm_brk(start, end - start);
 		if (BAD_ADDR(addr))
 			return addr;
 	}
@@ -514,9 +512,7 @@ static unsigned long load_elf_interp(struct elfhdr *interp_elf_ex,
 		elf_bss = ELF_PAGESTART(elf_bss + ELF_MIN_ALIGN - 1);
 
 		/* Map the last of the bss segment */
-		down_write(&current->mm->mmap_sem);
-		error = do_brk(elf_bss, last_bss - elf_bss);
-		up_write(&current->mm->mmap_sem);
+		error = vm_brk(elf_bss, last_bss - elf_bss);
 		if (BAD_ADDR(error))
 			goto out_close;
 	}
@@ -1072,11 +1068,8 @@ static int load_elf_library(struct file *file)
 	len = ELF_PAGESTART(eppnt->p_filesz + eppnt->p_vaddr +
 			    ELF_MIN_ALIGN - 1);
 	bss = eppnt->p_memsz + eppnt->p_vaddr;
-	if (bss > len) {
-		down_write(&current->mm->mmap_sem);
-		do_brk(len, bss - len);
-		up_write(&current->mm->mmap_sem);
-	}
+	if (bss > len)
+		vm_brk(len, bss - len);
 	error = 0;
 
 out_free_ph:
diff --git a/include/linux/mm.h b/include/linux/mm.h
index d8738a464b9..bfee4ad6680 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1415,7 +1415,8 @@ out:
 
 extern int do_munmap(struct mm_struct *, unsigned long, size_t);
 
-extern unsigned long do_brk(unsigned long, unsigned long);
+/* These take the mm semaphore themselves */
+extern unsigned long vm_brk(unsigned long, unsigned long);
 
 /* truncate.c */
 extern void truncate_inode_pages(struct address_space *, loff_t);
diff --git a/mm/mmap.c b/mm/mmap.c
index a7bf6a31c9f..df51891c864 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -240,6 +240,8 @@ static struct vm_area_struct *remove_vma(struct vm_area_struct *vma)
 	return next;
 }
 
+static unsigned long do_brk(unsigned long addr, unsigned long len);
+
 SYSCALL_DEFINE1(brk, unsigned long, brk)
 {
 	unsigned long rlim, retval;
@@ -2136,7 +2138,7 @@ static inline void verify_mm_writelocked(struct mm_struct *mm)
  *  anonymous maps.  eventually we may be able to do some
  *  brk-specific accounting here.
  */
-unsigned long do_brk(unsigned long addr, unsigned long len)
+static unsigned long do_brk(unsigned long addr, unsigned long len)
 {
 	struct mm_struct * mm = current->mm;
 	struct vm_area_struct * vma, * prev;
@@ -2232,7 +2234,17 @@ out:
 	return addr;
 }
 
-EXPORT_SYMBOL(do_brk);
+unsigned long vm_brk(unsigned long addr, unsigned long len)
+{
+	struct mm_struct *mm = current->mm;
+	unsigned long ret;
+
+	down_write(&mm->mmap_sem);
+	ret = do_brk(addr, len);
+	up_write(&mm->mmap_sem);
+	return ret;
+}
+EXPORT_SYMBOL(vm_brk);
 
 /* Release all mmaps. */
 void exit_mmap(struct mm_struct *mm)
diff --git a/mm/nommu.c b/mm/nommu.c
index f59e170fceb..634193324a6 100644
--- a/mm/nommu.c
+++ b/mm/nommu.c
@@ -1744,7 +1744,7 @@ void exit_mmap(struct mm_struct *mm)
 	kleave("");
 }
 
-unsigned long do_brk(unsigned long addr, unsigned long len)
+unsigned long vm_brk(unsigned long addr, unsigned long len)
 {
 	return -ENOMEM;
 }
-- 
cgit v1.2.3


From a46ef99d80817a167477ed1c8b4d90ee0c2e726f Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Fri, 20 Apr 2012 16:20:01 -0700
Subject: VM: add "vm_munmap()" helper function

Like the vm_brk() function, this is the same as "do_munmap()", except it
does the VM locking for the caller.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/ia64/kernel/perfmon.c       | 11 +++--------
 arch/sparc/kernel/sys_sparc_64.c |  7 +------
 arch/x86/kvm/x86.c               |  4 +---
 drivers/gpu/drm/i810/i810_dma.c  |  4 +---
 fs/aio.c                         |  7 ++-----
 include/linux/mm.h               |  1 +
 mm/mmap.c                        | 15 +++++++++------
 mm/nommu.c                       |  9 +++++++--
 8 files changed, 25 insertions(+), 33 deletions(-)

(limited to 'fs')

diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c
index 9d0fd7d5bb8..2777310b698 100644
--- a/arch/ia64/kernel/perfmon.c
+++ b/arch/ia64/kernel/perfmon.c
@@ -605,9 +605,9 @@ pfm_unprotect_ctx_ctxsw(pfm_context_t *x, unsigned long f)
 }
 
 static inline unsigned int
-pfm_do_munmap(struct mm_struct *mm, unsigned long addr, size_t len, int acct)
+pfm_vm_munmap(struct mm_struct *mm, unsigned long addr, size_t len)
 {
-	return do_munmap(mm, addr, len);
+	return vm_munmap(mm, addr, len);
 }
 
 static inline unsigned long 
@@ -1473,13 +1473,8 @@ pfm_remove_smpl_mapping(struct task_struct *task, void *vaddr, unsigned long siz
 	/*
 	 * does the actual unmapping
 	 */
-	down_write(&task->mm->mmap_sem);
-
-	DPRINT(("down_write done smpl_vaddr=%p size=%lu\n", vaddr, size));
+	r = pfm_vm_munmap(task->mm, (unsigned long)vaddr, size);
 
-	r = pfm_do_munmap(task->mm, (unsigned long)vaddr, size, 0);
-
-	up_write(&task->mm->mmap_sem);
 	if (r !=0) {
 		printk(KERN_ERR "perfmon: [%d] unable to unmap sampling buffer @%p size=%lu\n", task_pid_nr(task), vaddr, size);
 	}
diff --git a/arch/sparc/kernel/sys_sparc_64.c b/arch/sparc/kernel/sys_sparc_64.c
index 232df994953..022e57aadf5 100644
--- a/arch/sparc/kernel/sys_sparc_64.c
+++ b/arch/sparc/kernel/sys_sparc_64.c
@@ -566,15 +566,10 @@ out:
 
 SYSCALL_DEFINE2(64_munmap, unsigned long, addr, size_t, len)
 {
-	long ret;
-
 	if (invalid_64bit_range(addr, len))
 		return -EINVAL;
 
-	down_write(&current->mm->mmap_sem);
-	ret = do_munmap(current->mm, addr, len);
-	up_write(&current->mm->mmap_sem);
-	return ret;
+	return vm_munmap(current->mm, addr, len);
 }
 
 extern unsigned long do_mremap(unsigned long addr,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 4044ce0bf7c..8beb9ce7936 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -6366,10 +6366,8 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
 	if (!user_alloc && !old.user_alloc && old.rmap && !npages) {
 		int ret;
 
-		down_write(&current->mm->mmap_sem);
-		ret = do_munmap(current->mm, old.userspace_addr,
+		ret = vm_munmap(current->mm, old.userspace_addr,
 				old.npages * PAGE_SIZE);
-		up_write(&current->mm->mmap_sem);
 		if (ret < 0)
 			printk(KERN_WARNING
 			       "kvm_vm_ioctl_set_memory_region: "
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index 2c8a60c3b98..b85337f06fb 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -157,11 +157,9 @@ static int i810_unmap_buffer(struct drm_buf *buf)
 	if (buf_priv->currently_mapped != I810_BUF_MAPPED)
 		return -EINVAL;
 
-	down_write(&current->mm->mmap_sem);
-	retcode = do_munmap(current->mm,
+	retcode = vm_munmap(current->mm,
 			    (unsigned long)buf_priv->virtual,
 			    (size_t) buf->total);
-	up_write(&current->mm->mmap_sem);
 
 	buf_priv->currently_mapped = I810_BUF_UNMAPPED;
 	buf_priv->virtual = NULL;
diff --git a/fs/aio.c b/fs/aio.c
index da887604dfc..99bd790e8cd 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -92,11 +92,8 @@ static void aio_free_ring(struct kioctx *ctx)
 	for (i=0; i<info->nr_pages; i++)
 		put_page(info->ring_pages[i]);
 
-	if (info->mmap_size) {
-		down_write(&ctx->mm->mmap_sem);
-		do_munmap(ctx->mm, info->mmap_base, info->mmap_size);
-		up_write(&ctx->mm->mmap_sem);
-	}
+	if (info->mmap_size)
+		vm_munmap(ctx->mm, info->mmap_base, info->mmap_size);
 
 	if (info->ring_pages && info->ring_pages != info->internal_pages)
 		kfree(info->ring_pages);
diff --git a/include/linux/mm.h b/include/linux/mm.h
index bfee4ad6680..cb61950a3aa 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1417,6 +1417,7 @@ extern int do_munmap(struct mm_struct *, unsigned long, size_t);
 
 /* These take the mm semaphore themselves */
 extern unsigned long vm_brk(unsigned long, unsigned long);
+extern int vm_munmap(struct mm_struct *, unsigned long, size_t);
 
 /* truncate.c */
 extern void truncate_inode_pages(struct address_space *, loff_t);
diff --git a/mm/mmap.c b/mm/mmap.c
index df51891c864..4af45f519f1 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -2107,21 +2107,24 @@ int do_munmap(struct mm_struct *mm, unsigned long start, size_t len)
 
 	return 0;
 }
-
 EXPORT_SYMBOL(do_munmap);
 
-SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
+int vm_munmap(struct mm_struct *mm, unsigned long start, size_t len)
 {
 	int ret;
-	struct mm_struct *mm = current->mm;
-
-	profile_munmap(addr);
 
 	down_write(&mm->mmap_sem);
-	ret = do_munmap(mm, addr, len);
+	ret = do_munmap(mm, start, len);
 	up_write(&mm->mmap_sem);
 	return ret;
 }
+EXPORT_SYMBOL(vm_munmap);
+
+SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
+{
+	profile_munmap(addr);
+	return vm_munmap(current->mm, addr, len);
+}
 
 static inline void verify_mm_writelocked(struct mm_struct *mm)
 {
diff --git a/mm/nommu.c b/mm/nommu.c
index 634193324a6..11a69b22bd4 100644
--- a/mm/nommu.c
+++ b/mm/nommu.c
@@ -1709,16 +1709,21 @@ erase_whole_vma:
 }
 EXPORT_SYMBOL(do_munmap);
 
-SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
+int vm_munmap(struct mm_struct *mm, unsigned long addr, size_t len)
 {
 	int ret;
-	struct mm_struct *mm = current->mm;
 
 	down_write(&mm->mmap_sem);
 	ret = do_munmap(mm, addr, len);
 	up_write(&mm->mmap_sem);
 	return ret;
 }
+EXPORT_SYMBOL(vm_munmap);
+
+SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
+{
+	return vm_munmap(current->mm, addr, len);
+}
 
 /*
  * release all the mappings made in a process's VM space
-- 
cgit v1.2.3


From 6be5ceb02e98eaf6cfc4f8b12a896d04023f340d Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Fri, 20 Apr 2012 17:13:58 -0700
Subject: VM: add "vm_mmap()" helper function

This continues the theme started with vm_brk() and vm_munmap():
vm_mmap() does the same thing as do_mmap(), but additionally does the
required VM locking.

This uninlines (and rewrites it to be clearer) do_mmap(), which sadly
duplicates it in mm/mmap.c and mm/nommu.c.  But that way we don't have
to export our internal do_mmap_pgoff() function.

Some day we hopefully don't have to export do_mmap() either, if all
modular users can become the simpler vm_mmap() instead.  We're actually
very close to that already, with the notable exception of the (broken)
use in i810, and a couple of stragglers in binfmt_elf.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/tile/kernel/single_step.c          |  4 +---
 arch/x86/ia32/ia32_aout.c               | 12 +++---------
 arch/x86/kvm/x86.c                      |  4 +---
 drivers/gpu/drm/drm_bufs.c              | 12 ++++--------
 drivers/gpu/drm/exynos/exynos_drm_gem.c |  4 +---
 drivers/gpu/drm/i810/i810_dma.c         |  1 +
 drivers/gpu/drm/i915/i915_gem.c         |  4 +---
 fs/binfmt_aout.c                        | 12 +++---------
 fs/binfmt_elf.c                         |  8 ++------
 fs/binfmt_elf_fdpic.c                   | 18 ++++--------------
 fs/binfmt_flat.c                        | 12 +++---------
 fs/binfmt_som.c                         | 12 +++---------
 include/linux/mm.h                      | 23 ++++++-----------------
 mm/mmap.c                               | 29 +++++++++++++++++++++++++++--
 mm/nommu.c                              | 29 +++++++++++++++++++++++++++--
 15 files changed, 87 insertions(+), 97 deletions(-)

(limited to 'fs')

diff --git a/arch/tile/kernel/single_step.c b/arch/tile/kernel/single_step.c
index 9efbc1391b3..89529c9f060 100644
--- a/arch/tile/kernel/single_step.c
+++ b/arch/tile/kernel/single_step.c
@@ -346,12 +346,10 @@ void single_step_once(struct pt_regs *regs)
 		}
 
 		/* allocate a cache line of writable, executable memory */
-		down_write(&current->mm->mmap_sem);
-		buffer = (void __user *) do_mmap(NULL, 0, 64,
+		buffer = (void __user *) vm_mmap(NULL, 0, 64,
 					  PROT_EXEC | PROT_READ | PROT_WRITE,
 					  MAP_PRIVATE | MAP_ANONYMOUS,
 					  0);
-		up_write(&current->mm->mmap_sem);
 
 		if (IS_ERR((void __force *)buffer)) {
 			kfree(state);
diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c
index b6817ee9033..4824fb45560 100644
--- a/arch/x86/ia32/ia32_aout.c
+++ b/arch/x86/ia32/ia32_aout.c
@@ -379,26 +379,22 @@ static int load_aout_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 			goto beyond_if;
 		}
 
-		down_write(&current->mm->mmap_sem);
-		error = do_mmap(bprm->file, N_TXTADDR(ex), ex.a_text,
+		error = vm_mmap(bprm->file, N_TXTADDR(ex), ex.a_text,
 				PROT_READ | PROT_EXEC,
 				MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE |
 				MAP_EXECUTABLE | MAP_32BIT,
 				fd_offset);
-		up_write(&current->mm->mmap_sem);
 
 		if (error != N_TXTADDR(ex)) {
 			send_sig(SIGKILL, current, 0);
 			return error;
 		}
 
-		down_write(&current->mm->mmap_sem);
-		error = do_mmap(bprm->file, N_DATADDR(ex), ex.a_data,
+		error = vm_mmap(bprm->file, N_DATADDR(ex), ex.a_data,
 				PROT_READ | PROT_WRITE | PROT_EXEC,
 				MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE |
 				MAP_EXECUTABLE | MAP_32BIT,
 				fd_offset + ex.a_text);
-		up_write(&current->mm->mmap_sem);
 		if (error != N_DATADDR(ex)) {
 			send_sig(SIGKILL, current, 0);
 			return error;
@@ -482,12 +478,10 @@ static int load_aout_library(struct file *file)
 		goto out;
 	}
 	/* Now use mmap to map the library into memory. */
-	down_write(&current->mm->mmap_sem);
-	error = do_mmap(file, start_addr, ex.a_text + ex.a_data,
+	error = vm_mmap(file, start_addr, ex.a_text + ex.a_data,
 			PROT_READ | PROT_WRITE | PROT_EXEC,
 			MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE | MAP_32BIT,
 			N_TXTOFF(ex));
-	up_write(&current->mm->mmap_sem);
 	retval = error;
 	if (error != start_addr)
 		goto out;
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 8beb9ce7936..1457be305fb 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -6336,13 +6336,11 @@ int kvm_arch_prepare_memory_region(struct kvm *kvm,
 		if (npages && !old.rmap) {
 			unsigned long userspace_addr;
 
-			down_write(&current->mm->mmap_sem);
-			userspace_addr = do_mmap(NULL, 0,
+			userspace_addr = vm_mmap(NULL, 0,
 						 npages * PAGE_SIZE,
 						 PROT_READ | PROT_WRITE,
 						 map_flags,
 						 0);
-			up_write(&current->mm->mmap_sem);
 
 			if (IS_ERR((void *)userspace_addr))
 				return PTR_ERR((void *)userspace_addr);
diff --git a/drivers/gpu/drm/drm_bufs.c b/drivers/gpu/drm/drm_bufs.c
index 30372f7b2d4..348b367debe 100644
--- a/drivers/gpu/drm/drm_bufs.c
+++ b/drivers/gpu/drm/drm_bufs.c
@@ -1510,8 +1510,8 @@ int drm_freebufs(struct drm_device *dev, void *data,
  * \param arg pointer to a drm_buf_map structure.
  * \return zero on success or a negative number on failure.
  *
- * Maps the AGP, SG or PCI buffer region with do_mmap(), and copies information
- * about each buffer into user space. For PCI buffers, it calls do_mmap() with
+ * Maps the AGP, SG or PCI buffer region with vm_mmap(), and copies information
+ * about each buffer into user space. For PCI buffers, it calls vm_mmap() with
  * offset equal to 0, which drm_mmap() interpretes as PCI buffers and calls
  * drm_mmap_dma().
  */
@@ -1553,18 +1553,14 @@ int drm_mapbufs(struct drm_device *dev, void *data,
 				retcode = -EINVAL;
 				goto done;
 			}
-			down_write(&current->mm->mmap_sem);
-			virtual = do_mmap(file_priv->filp, 0, map->size,
+			virtual = vm_mmap(file_priv->filp, 0, map->size,
 					  PROT_READ | PROT_WRITE,
 					  MAP_SHARED,
 					  token);
-			up_write(&current->mm->mmap_sem);
 		} else {
-			down_write(&current->mm->mmap_sem);
-			virtual = do_mmap(file_priv->filp, 0, dma->byte_count,
+			virtual = vm_mmap(file_priv->filp, 0, dma->byte_count,
 					  PROT_READ | PROT_WRITE,
 					  MAP_SHARED, 0);
-			up_write(&current->mm->mmap_sem);
 		}
 		if (virtual > -1024UL) {
 			/* Real error */
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index 26d51979116..392ce71ed6a 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -581,10 +581,8 @@ int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
 	obj->filp->f_op = &exynos_drm_gem_fops;
 	obj->filp->private_data = obj;
 
-	down_write(&current->mm->mmap_sem);
-	addr = do_mmap(obj->filp, 0, args->size,
+	addr = vm_mmap(obj->filp, 0, args->size,
 			PROT_READ | PROT_WRITE, MAP_SHARED, 0);
-	up_write(&current->mm->mmap_sem);
 
 	drm_gem_object_unreference_unlocked(obj);
 
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index b85337f06fb..a4ba453b3d2 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -129,6 +129,7 @@ static int i810_map_buffer(struct drm_buf *buf, struct drm_file *file_priv)
 	if (buf_priv->currently_mapped == I810_BUF_MAPPED)
 		return -EINVAL;
 
+	/* This is all entirely broken */
 	down_write(&current->mm->mmap_sem);
 	old_fops = file_priv->filp->f_op;
 	file_priv->filp->f_op = &i810_buffer_fops;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 0e3c6acde95..0d1e4b7b4b9 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1087,11 +1087,9 @@ i915_gem_mmap_ioctl(struct drm_device *dev, void *data,
 	if (obj == NULL)
 		return -ENOENT;
 
-	down_write(&current->mm->mmap_sem);
-	addr = do_mmap(obj->filp, 0, args->size,
+	addr = vm_mmap(obj->filp, 0, args->size,
 		       PROT_READ | PROT_WRITE, MAP_SHARED,
 		       args->offset);
-	up_write(&current->mm->mmap_sem);
 	drm_gem_object_unreference_unlocked(obj);
 	if (IS_ERR((void *)addr))
 		return addr;
diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c
index 88527492b91..d146e181d10 100644
--- a/fs/binfmt_aout.c
+++ b/fs/binfmt_aout.c
@@ -319,24 +319,20 @@ static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs)
 			goto beyond_if;
 		}
 
-		down_write(&current->mm->mmap_sem);
-		error = do_mmap(bprm->file, N_TXTADDR(ex), ex.a_text,
+		error = vm_mmap(bprm->file, N_TXTADDR(ex), ex.a_text,
 			PROT_READ | PROT_EXEC,
 			MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE | MAP_EXECUTABLE,
 			fd_offset);
-		up_write(&current->mm->mmap_sem);
 
 		if (error != N_TXTADDR(ex)) {
 			send_sig(SIGKILL, current, 0);
 			return error;
 		}
 
-		down_write(&current->mm->mmap_sem);
- 		error = do_mmap(bprm->file, N_DATADDR(ex), ex.a_data,
+		error = vm_mmap(bprm->file, N_DATADDR(ex), ex.a_data,
 				PROT_READ | PROT_WRITE | PROT_EXEC,
 				MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE | MAP_EXECUTABLE,
 				fd_offset + ex.a_text);
-		up_write(&current->mm->mmap_sem);
 		if (error != N_DATADDR(ex)) {
 			send_sig(SIGKILL, current, 0);
 			return error;
@@ -417,12 +413,10 @@ static int load_aout_library(struct file *file)
 		goto out;
 	}
 	/* Now use mmap to map the library into memory. */
-	down_write(&current->mm->mmap_sem);
-	error = do_mmap(file, start_addr, ex.a_text + ex.a_data,
+	error = vm_mmap(file, start_addr, ex.a_text + ex.a_data,
 			PROT_READ | PROT_WRITE | PROT_EXEC,
 			MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE,
 			N_TXTOFF(ex));
-	up_write(&current->mm->mmap_sem);
 	retval = error;
 	if (error != start_addr)
 		goto out;
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 0708a0bf0ba..16f73541707 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -958,10 +958,8 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs)
 		   and some applications "depend" upon this behavior.
 		   Since we do not have the power to recompile these, we
 		   emulate the SVr4 behavior. Sigh. */
-		down_write(&current->mm->mmap_sem);
-		error = do_mmap(NULL, 0, PAGE_SIZE, PROT_READ | PROT_EXEC,
+		error = vm_mmap(NULL, 0, PAGE_SIZE, PROT_READ | PROT_EXEC,
 				MAP_FIXED | MAP_PRIVATE, 0);
-		up_write(&current->mm->mmap_sem);
 	}
 
 #ifdef ELF_PLAT_INIT
@@ -1046,8 +1044,7 @@ static int load_elf_library(struct file *file)
 		eppnt++;
 
 	/* Now use mmap to map the library into memory. */
-	down_write(&current->mm->mmap_sem);
-	error = do_mmap(file,
+	error = vm_mmap(file,
 			ELF_PAGESTART(eppnt->p_vaddr),
 			(eppnt->p_filesz +
 			 ELF_PAGEOFFSET(eppnt->p_vaddr)),
@@ -1055,7 +1052,6 @@ static int load_elf_library(struct file *file)
 			MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE,
 			(eppnt->p_offset -
 			 ELF_PAGEOFFSET(eppnt->p_vaddr)));
-	up_write(&current->mm->mmap_sem);
 	if (error != ELF_PAGESTART(eppnt->p_vaddr))
 		goto out_free_ph;
 
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index 9bd5612a822..d390a0fffc6 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -390,21 +390,17 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm,
 	    (executable_stack == EXSTACK_DEFAULT && VM_STACK_FLAGS & VM_EXEC))
 		stack_prot |= PROT_EXEC;
 
-	down_write(&current->mm->mmap_sem);
-	current->mm->start_brk = do_mmap(NULL, 0, stack_size, stack_prot,
+	current->mm->start_brk = vm_mmap(NULL, 0, stack_size, stack_prot,
 					 MAP_PRIVATE | MAP_ANONYMOUS |
 					 MAP_UNINITIALIZED | MAP_GROWSDOWN,
 					 0);
 
 	if (IS_ERR_VALUE(current->mm->start_brk)) {
-		up_write(&current->mm->mmap_sem);
 		retval = current->mm->start_brk;
 		current->mm->start_brk = 0;
 		goto error_kill;
 	}
 
-	up_write(&current->mm->mmap_sem);
-
 	current->mm->brk = current->mm->start_brk;
 	current->mm->context.end_brk = current->mm->start_brk;
 	current->mm->context.end_brk +=
@@ -955,10 +951,8 @@ static int elf_fdpic_map_file_constdisp_on_uclinux(
 	if (params->flags & ELF_FDPIC_FLAG_EXECUTABLE)
 		mflags |= MAP_EXECUTABLE;
 
-	down_write(&mm->mmap_sem);
-	maddr = do_mmap(NULL, load_addr, top - base,
+	maddr = vm_mmap(NULL, load_addr, top - base,
 			PROT_READ | PROT_WRITE | PROT_EXEC, mflags, 0);
-	up_write(&mm->mmap_sem);
 	if (IS_ERR_VALUE(maddr))
 		return (int) maddr;
 
@@ -1096,10 +1090,8 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params,
 
 		/* create the mapping */
 		disp = phdr->p_vaddr & ~PAGE_MASK;
-		down_write(&mm->mmap_sem);
-		maddr = do_mmap(file, maddr, phdr->p_memsz + disp, prot, flags,
+		maddr = vm_mmap(file, maddr, phdr->p_memsz + disp, prot, flags,
 				phdr->p_offset - disp);
-		up_write(&mm->mmap_sem);
 
 		kdebug("mmap[%d] <file> sz=%lx pr=%x fl=%x of=%lx --> %08lx",
 		       loop, phdr->p_memsz + disp, prot, flags,
@@ -1143,10 +1135,8 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params,
 			unsigned long xmaddr;
 
 			flags |= MAP_FIXED | MAP_ANONYMOUS;
-			down_write(&mm->mmap_sem);
-			xmaddr = do_mmap(NULL, xaddr, excess - excess1,
+			xmaddr = vm_mmap(NULL, xaddr, excess - excess1,
 					 prot, flags, 0);
-			up_write(&mm->mmap_sem);
 
 			kdebug("mmap[%d] <anon>"
 			       " ad=%lx sz=%lx pr=%x fl=%x of=0 --> %08lx",
diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c
index 024d20ee3ca..6b2daf99fab 100644
--- a/fs/binfmt_flat.c
+++ b/fs/binfmt_flat.c
@@ -542,10 +542,8 @@ static int load_flat_file(struct linux_binprm * bprm,
 		 */
 		DBG_FLT("BINFMT_FLAT: ROM mapping of file (we hope)\n");
 
-		down_write(&current->mm->mmap_sem);
-		textpos = do_mmap(bprm->file, 0, text_len, PROT_READ|PROT_EXEC,
+		textpos = vm_mmap(bprm->file, 0, text_len, PROT_READ|PROT_EXEC,
 				  MAP_PRIVATE|MAP_EXECUTABLE, 0);
-		up_write(&current->mm->mmap_sem);
 		if (!textpos || IS_ERR_VALUE(textpos)) {
 			if (!textpos)
 				textpos = (unsigned long) -ENOMEM;
@@ -556,10 +554,8 @@ static int load_flat_file(struct linux_binprm * bprm,
 
 		len = data_len + extra + MAX_SHARED_LIBS * sizeof(unsigned long);
 		len = PAGE_ALIGN(len);
-		down_write(&current->mm->mmap_sem);
-		realdatastart = do_mmap(0, 0, len,
+		realdatastart = vm_mmap(0, 0, len,
 			PROT_READ|PROT_WRITE|PROT_EXEC, MAP_PRIVATE, 0);
-		up_write(&current->mm->mmap_sem);
 
 		if (realdatastart == 0 || IS_ERR_VALUE(realdatastart)) {
 			if (!realdatastart)
@@ -603,10 +599,8 @@ static int load_flat_file(struct linux_binprm * bprm,
 
 		len = text_len + data_len + extra + MAX_SHARED_LIBS * sizeof(unsigned long);
 		len = PAGE_ALIGN(len);
-		down_write(&current->mm->mmap_sem);
-		textpos = do_mmap(0, 0, len,
+		textpos = vm_mmap(0, 0, len,
 			PROT_READ | PROT_EXEC | PROT_WRITE, MAP_PRIVATE, 0);
-		up_write(&current->mm->mmap_sem);
 
 		if (!textpos || IS_ERR_VALUE(textpos)) {
 			if (!textpos)
diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c
index e4fc746629a..4517aaff61b 100644
--- a/fs/binfmt_som.c
+++ b/fs/binfmt_som.c
@@ -147,10 +147,8 @@ static int map_som_binary(struct file *file,
 	code_size = SOM_PAGEALIGN(hpuxhdr->exec_tsize);
 	current->mm->start_code = code_start;
 	current->mm->end_code = code_start + code_size;
-	down_write(&current->mm->mmap_sem);
-	retval = do_mmap(file, code_start, code_size, prot,
+	retval = vm_mmap(file, code_start, code_size, prot,
 			flags, SOM_PAGESTART(hpuxhdr->exec_tfile));
-	up_write(&current->mm->mmap_sem);
 	if (retval < 0 && retval > -1024)
 		goto out;
 
@@ -158,20 +156,16 @@ static int map_som_binary(struct file *file,
 	data_size = SOM_PAGEALIGN(hpuxhdr->exec_dsize);
 	current->mm->start_data = data_start;
 	current->mm->end_data = bss_start = data_start + data_size;
-	down_write(&current->mm->mmap_sem);
-	retval = do_mmap(file, data_start, data_size,
+	retval = vm_mmap(file, data_start, data_size,
 			prot | PROT_WRITE, flags,
 			SOM_PAGESTART(hpuxhdr->exec_dfile));
-	up_write(&current->mm->mmap_sem);
 	if (retval < 0 && retval > -1024)
 		goto out;
 
 	som_brk = bss_start + SOM_PAGEALIGN(hpuxhdr->exec_bsize);
 	current->mm->start_brk = current->mm->brk = som_brk;
-	down_write(&current->mm->mmap_sem);
-	retval = do_mmap(NULL, bss_start, som_brk - bss_start,
+	retval = vm_mmap(NULL, bss_start, som_brk - bss_start,
 			prot | PROT_WRITE, MAP_FIXED | MAP_PRIVATE, 0);
-	up_write(&current->mm->mmap_sem);
 	if (retval > 0 || retval < -1024)
 		retval = 0;
 out:
diff --git a/include/linux/mm.h b/include/linux/mm.h
index cb61950a3aa..86a692c3b23 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1393,31 +1393,20 @@ extern int install_special_mapping(struct mm_struct *mm,
 
 extern unsigned long get_unmapped_area(struct file *, unsigned long, unsigned long, unsigned long, unsigned long);
 
-extern unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
-	unsigned long len, unsigned long prot,
-	unsigned long flag, unsigned long pgoff);
 extern unsigned long mmap_region(struct file *file, unsigned long addr,
 	unsigned long len, unsigned long flags,
 	vm_flags_t vm_flags, unsigned long pgoff);
-
-static inline unsigned long do_mmap(struct file *file, unsigned long addr,
-	unsigned long len, unsigned long prot,
-	unsigned long flag, unsigned long offset)
-{
-	unsigned long ret = -EINVAL;
-	if ((offset + PAGE_ALIGN(len)) < offset)
-		goto out;
-	if (!(offset & ~PAGE_MASK))
-		ret = do_mmap_pgoff(file, addr, len, prot, flag, offset >> PAGE_SHIFT);
-out:
-	return ret;
-}
-
+extern unsigned long do_mmap(struct file *, unsigned long,
+        unsigned long, unsigned long,
+        unsigned long, unsigned long);
 extern int do_munmap(struct mm_struct *, unsigned long, size_t);
 
 /* These take the mm semaphore themselves */
 extern unsigned long vm_brk(unsigned long, unsigned long);
 extern int vm_munmap(struct mm_struct *, unsigned long, size_t);
+extern unsigned long vm_mmap(struct file *, unsigned long,
+        unsigned long, unsigned long,
+        unsigned long, unsigned long);
 
 /* truncate.c */
 extern void truncate_inode_pages(struct address_space *, loff_t);
diff --git a/mm/mmap.c b/mm/mmap.c
index 4af45f519f1..b38b47ef1f7 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -953,7 +953,7 @@ static inline unsigned long round_hint_to_min(unsigned long hint)
  * The caller must hold down_write(&current->mm->mmap_sem).
  */
 
-unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
+static unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
 			unsigned long len, unsigned long prot,
 			unsigned long flags, unsigned long pgoff)
 {
@@ -1089,7 +1089,32 @@ unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
 
 	return mmap_region(file, addr, len, flags, vm_flags, pgoff);
 }
-EXPORT_SYMBOL(do_mmap_pgoff);
+
+unsigned long do_mmap(struct file *file, unsigned long addr,
+	unsigned long len, unsigned long prot,
+	unsigned long flag, unsigned long offset)
+{
+	if (unlikely(offset + PAGE_ALIGN(len) < offset))
+		return -EINVAL;
+	if (unlikely(offset & ~PAGE_MASK))
+		return -EINVAL;
+	return do_mmap_pgoff(file, addr, len, prot, flag, offset >> PAGE_SHIFT);
+}
+EXPORT_SYMBOL(do_mmap);
+
+unsigned long vm_mmap(struct file *file, unsigned long addr,
+	unsigned long len, unsigned long prot,
+	unsigned long flag, unsigned long offset)
+{
+	unsigned long ret;
+	struct mm_struct *mm = current->mm;
+
+	down_write(&mm->mmap_sem);
+	ret = do_mmap(file, addr, len, prot, flag, offset);
+	up_write(&mm->mmap_sem);
+	return ret;
+}
+EXPORT_SYMBOL(vm_mmap);
 
 SYSCALL_DEFINE6(mmap_pgoff, unsigned long, addr, unsigned long, len,
 		unsigned long, prot, unsigned long, flags,
diff --git a/mm/nommu.c b/mm/nommu.c
index 11a69b22bd4..dd00383be2d 100644
--- a/mm/nommu.c
+++ b/mm/nommu.c
@@ -1233,7 +1233,7 @@ enomem:
 /*
  * handle mapping creation for uClinux
  */
-unsigned long do_mmap_pgoff(struct file *file,
+static unsigned long do_mmap_pgoff(struct file *file,
 			    unsigned long addr,
 			    unsigned long len,
 			    unsigned long prot,
@@ -1470,7 +1470,32 @@ error_getting_region:
 	show_free_areas(0);
 	return -ENOMEM;
 }
-EXPORT_SYMBOL(do_mmap_pgoff);
+
+unsigned long do_mmap(struct file *file, unsigned long addr,
+	unsigned long len, unsigned long prot,
+	unsigned long flag, unsigned long offset)
+{
+	if (unlikely(offset + PAGE_ALIGN(len) < offset))
+		return -EINVAL;
+	if (unlikely(offset & ~PAGE_MASK))
+		return -EINVAL;
+	return do_mmap_pgoff(file, addr, len, prot, flag, offset >> PAGE_SHIFT);
+}
+EXPORT_SYMBOL(do_mmap);
+
+unsigned long vm_mmap(struct file *file, unsigned long addr,
+	unsigned long len, unsigned long prot,
+	unsigned long flag, unsigned long offset)
+{
+	unsigned long ret;
+	struct mm_struct *mm = current->mm;
+
+	down_write(&mm->mmap_sem);
+	ret = do_mmap(file, addr, len, prot, flag, offset);
+	up_write(&mm->mmap_sem);
+	return ret;
+}
+EXPORT_SYMBOL(vm_mmap);
 
 SYSCALL_DEFINE6(mmap_pgoff, unsigned long, addr, unsigned long, len,
 		unsigned long, prot, unsigned long, flags,
-- 
cgit v1.2.3


From 936af1576e4c24b466380fc2b8d93352161d13b0 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 20 Apr 2012 21:49:41 -0400
Subject: aio: don't bother with unmapping when aio_free_ring() is coming from
 exit_aio()

... since exit_mmap() is coming and it will munmap() everything anyway.
In all other cases aio_free_ring() has ctx->mm == current->mm; moreover,
all other callers of vm_munmap() have mm == current->mm, so this will
allow us to get rid of mm argument of vm_munmap().

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 fs/aio.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

(limited to 'fs')

diff --git a/fs/aio.c b/fs/aio.c
index 99bd790e8cd..976e33d9741 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -92,8 +92,10 @@ static void aio_free_ring(struct kioctx *ctx)
 	for (i=0; i<info->nr_pages; i++)
 		put_page(info->ring_pages[i]);
 
-	if (info->mmap_size)
+	if (info->mmap_size) {
+		BUG_ON(ctx->mm != current->mm);
 		vm_munmap(ctx->mm, info->mmap_base, info->mmap_size);
+	}
 
 	if (info->ring_pages && info->ring_pages != info->internal_pages)
 		kfree(info->ring_pages);
@@ -386,6 +388,17 @@ void exit_aio(struct mm_struct *mm)
 				"exit_aio:ioctx still alive: %d %d %d\n",
 				atomic_read(&ctx->users), ctx->dead,
 				ctx->reqs_active);
+		/*
+		 * We don't need to bother with munmap() here -
+		 * exit_mmap(mm) is coming and it'll unmap everything.
+		 * Since aio_free_ring() uses non-zero ->mmap_size
+		 * as indicator that it needs to unmap the area,
+		 * just set it to 0; aio_free_ring() is the only
+		 * place that uses ->mmap_size, so it's safe.
+		 * That way we get all munmap done to current->mm -
+		 * all other callers have ctx->mm == current->mm.
+		 */
+		ctx->ring_info.mmap_size = 0;
 		put_ioctx(ctx);
 	}
 }
-- 
cgit v1.2.3


From bfce281c287a427d0841fadf5d59242757b4e620 Mon Sep 17 00:00:00 2001
From: Al Viro <viro@zeniv.linux.org.uk>
Date: Fri, 20 Apr 2012 21:57:04 -0400
Subject: kill mm argument of vm_munmap()

it's always current->mm

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
 arch/ia64/kernel/perfmon.c       | 2 +-
 arch/sparc/kernel/sys_sparc_64.c | 2 +-
 arch/x86/kvm/x86.c               | 2 +-
 drivers/gpu/drm/i810/i810_dma.c  | 3 +--
 fs/aio.c                         | 2 +-
 include/linux/mm.h               | 2 +-
 mm/mmap.c                        | 5 +++--
 mm/nommu.c                       | 5 +++--
 8 files changed, 12 insertions(+), 11 deletions(-)

(limited to 'fs')

diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c
index 899c0fa5b49..f00ba025375 100644
--- a/arch/ia64/kernel/perfmon.c
+++ b/arch/ia64/kernel/perfmon.c
@@ -1468,7 +1468,7 @@ pfm_remove_smpl_mapping(void *vaddr, unsigned long size)
 	/*
 	 * does the actual unmapping
 	 */
-	r = vm_munmap(current->mm, (unsigned long)vaddr, size);
+	r = vm_munmap((unsigned long)vaddr, size);
 
 	if (r !=0) {
 		printk(KERN_ERR "perfmon: [%d] unable to unmap sampling buffer @%p size=%lu\n", task_pid_nr(task), vaddr, size);
diff --git a/arch/sparc/kernel/sys_sparc_64.c b/arch/sparc/kernel/sys_sparc_64.c
index 022e57aadf5..3ee51f189a5 100644
--- a/arch/sparc/kernel/sys_sparc_64.c
+++ b/arch/sparc/kernel/sys_sparc_64.c
@@ -569,7 +569,7 @@ SYSCALL_DEFINE2(64_munmap, unsigned long, addr, size_t, len)
 	if (invalid_64bit_range(addr, len))
 		return -EINVAL;
 
-	return vm_munmap(current->mm, addr, len);
+	return vm_munmap(addr, len);
 }
 
 extern unsigned long do_mremap(unsigned long addr,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 1457be305fb..91a5e989abc 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -6364,7 +6364,7 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
 	if (!user_alloc && !old.user_alloc && old.rmap && !npages) {
 		int ret;
 
-		ret = vm_munmap(current->mm, old.userspace_addr,
+		ret = vm_munmap(old.userspace_addr,
 				old.npages * PAGE_SIZE);
 		if (ret < 0)
 			printk(KERN_WARNING
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index a4ba453b3d2..f920fb5e42b 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -158,8 +158,7 @@ static int i810_unmap_buffer(struct drm_buf *buf)
 	if (buf_priv->currently_mapped != I810_BUF_MAPPED)
 		return -EINVAL;
 
-	retcode = vm_munmap(current->mm,
-			    (unsigned long)buf_priv->virtual,
+	retcode = vm_munmap((unsigned long)buf_priv->virtual,
 			    (size_t) buf->total);
 
 	buf_priv->currently_mapped = I810_BUF_UNMAPPED;
diff --git a/fs/aio.c b/fs/aio.c
index 976e33d9741..67a6db3e1b6 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -94,7 +94,7 @@ static void aio_free_ring(struct kioctx *ctx)
 
 	if (info->mmap_size) {
 		BUG_ON(ctx->mm != current->mm);
-		vm_munmap(ctx->mm, info->mmap_base, info->mmap_size);
+		vm_munmap(info->mmap_base, info->mmap_size);
 	}
 
 	if (info->ring_pages && info->ring_pages != info->internal_pages)
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 86a692c3b23..74aa71bea1e 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1403,7 +1403,7 @@ extern int do_munmap(struct mm_struct *, unsigned long, size_t);
 
 /* These take the mm semaphore themselves */
 extern unsigned long vm_brk(unsigned long, unsigned long);
-extern int vm_munmap(struct mm_struct *, unsigned long, size_t);
+extern int vm_munmap(unsigned long, size_t);
 extern unsigned long vm_mmap(struct file *, unsigned long,
         unsigned long, unsigned long,
         unsigned long, unsigned long);
diff --git a/mm/mmap.c b/mm/mmap.c
index b38b47ef1f7..848ef52d960 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -2134,9 +2134,10 @@ int do_munmap(struct mm_struct *mm, unsigned long start, size_t len)
 }
 EXPORT_SYMBOL(do_munmap);
 
-int vm_munmap(struct mm_struct *mm, unsigned long start, size_t len)
+int vm_munmap(unsigned long start, size_t len)
 {
 	int ret;
+	struct mm_struct *mm = current->mm;
 
 	down_write(&mm->mmap_sem);
 	ret = do_munmap(mm, start, len);
@@ -2148,7 +2149,7 @@ EXPORT_SYMBOL(vm_munmap);
 SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
 {
 	profile_munmap(addr);
-	return vm_munmap(current->mm, addr, len);
+	return vm_munmap(addr, len);
 }
 
 static inline void verify_mm_writelocked(struct mm_struct *mm)
diff --git a/mm/nommu.c b/mm/nommu.c
index dd00383be2d..bb8f4f004a8 100644
--- a/mm/nommu.c
+++ b/mm/nommu.c
@@ -1734,8 +1734,9 @@ erase_whole_vma:
 }
 EXPORT_SYMBOL(do_munmap);
 
-int vm_munmap(struct mm_struct *mm, unsigned long addr, size_t len)
+int vm_munmap(unsigned long addr, size_t len)
 {
+	struct mm_struct *mm = current->mm;
 	int ret;
 
 	down_write(&mm->mmap_sem);
@@ -1747,7 +1748,7 @@ EXPORT_SYMBOL(vm_munmap);
 
 SYSCALL_DEFINE2(munmap, unsigned long, addr, size_t, len)
 {
-	return vm_munmap(current->mm, addr, len);
+	return vm_munmap(addr, len);
 }
 
 /*
-- 
cgit v1.2.3