Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs: (28 commits)
  net/9p: Return error on read with NULL buffer
  9p: Add datasync to client side TFSYNC/RFSYNC for dotl
  net/9p: Return error if we fail to encode protocol data
  fs/9p: Use generic_file_open with lookup_instantiate_filp
  fs/9p: Add missing iput in v9fs_vfs_lookup
  fs/9p: Use mknod 9p operation on create without open request
  net/9p: Add waitq to VirtIO transport.
  [net/9p]Serialize virtqueue operations to make VirtIO transport SMP safe.
  9p: Implement TREADLINK operation for 9p2000.L
  9p: Use V9FS_MAGIC in statfs
  9p: Implement TGETLOCK
  9p: Implement TLOCK
  [9p] Introduce client side TFSYNC/RFSYNC for dotl.
  [fs/9p] Add file_operations for cached mode in dotl protocol.
  fs/9p: Add access = client option to opt in acl evaluation.
  fs/9p: Implement create time inheritance
  fs/9p: Update ACL on chmod
  fs/9p: Implement setting posix acl
  fs/9p: Add xattr callbacks for POSIX ACL
  fs/9p: Implement POSIX ACL permission checking function
  ...

Documentation/filesystems/9p.txt

@@ -111,7 +111,7 @@ OPTIONS
   		This can be used to share devices/named pipes/sockets between
 		hosts.  This functionality will be expanded in later versions.
 
-  access	there are three access modes.
+  access	there are four access modes.
 			user  = if a user tries to access a file on v9fs
 			        filesystem for the first time, v9fs sends an
 			        attach command (Tattach) for that user.
@@ -120,6 +120,8 @@ OPTIONS
 				the files on the mounted filesystem
 			any   = v9fs does single attach and performs all
 				operations as one user
+			client = ACL based access check on the 9p client
+			         side for access validation
 
   cachetag	cache tag to use the specified persistent cache.
 		cache tags for existing cache sessions can be listed at

fs/9p/Kconfig

@@ -17,3 +17,16 @@ config 9P_FSCACHE
 	  Choose Y here to enable persistent, read-only local
 	  caching support for 9p clients using FS-Cache
 
+
+config 9P_FS_POSIX_ACL
+	bool "9P POSIX Access Control Lists"
+	depends on 9P_FS
+	select FS_POSIX_ACL
+	help
+	  POSIX Access Control Lists (ACLs) support permissions for users and
+	  groups beyond the owner/group/world scheme.
+
+	  To learn more about Access Control Lists, visit the POSIX ACLs for
+	  Linux website <http://acl.bestbits.at/>.
+
+	  If you don't know what Access Control Lists are, say N

fs/9p/Makefile

@@ -13,3 +13,4 @@ obj-$(CONFIG_9P_FS) := 9p.o
 	xattr_user.o
 
 9p-$(CONFIG_9P_FSCACHE) += cache.o
+9p-$(CONFIG_9P_FS_POSIX_ACL) += acl.o

fs/9p/acl.c

@@ -0,0 +1,392 @@
+/*
+ * Copyright IBM Corporation, 2010
+ * Author Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <net/9p/9p.h>
+#include <net/9p/client.h>
+#include <linux/slab.h>
+#include <linux/sched.h>
+#include <linux/posix_acl_xattr.h>
+#include "xattr.h"
+#include "acl.h"
+#include "v9fs_vfs.h"
+#include "v9fs.h"
+
+static struct posix_acl *__v9fs_get_acl(struct p9_fid *fid, char *name)
+{
+	ssize_t size;
+	void *value = NULL;
+	struct posix_acl *acl = NULL;;
+
+	size = v9fs_fid_xattr_get(fid, name, NULL, 0);
+	if (size > 0) {
+		value = kzalloc(size, GFP_NOFS);
+		if (!value)
+			return ERR_PTR(-ENOMEM);
+		size = v9fs_fid_xattr_get(fid, name, value, size);
+		if (size > 0) {
+			acl = posix_acl_from_xattr(value, size);
+			if (IS_ERR(acl))
+				goto err_out;
+		}
+	} else if (size == -ENODATA || size == 0 ||
+		   size == -ENOSYS || size == -EOPNOTSUPP) {
+		acl = NULL;
+	} else
+		acl = ERR_PTR(-EIO);
+
+err_out:
+	kfree(value);
+	return acl;
+}
+
+int v9fs_get_acl(struct inode *inode, struct p9_fid *fid)
+{
+	int retval = 0;
+	struct posix_acl *pacl, *dacl;
+	struct v9fs_session_info *v9ses;
+
+	v9ses = v9fs_inode2v9ses(inode);
+	if ((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT) {
+		set_cached_acl(inode, ACL_TYPE_DEFAULT, NULL);
+		set_cached_acl(inode, ACL_TYPE_ACCESS, NULL);
+		return 0;
+	}
+	/* get the default/access acl values and cache them */
+	dacl = __v9fs_get_acl(fid, POSIX_ACL_XATTR_DEFAULT);
+	pacl = __v9fs_get_acl(fid, POSIX_ACL_XATTR_ACCESS);
+
+	if (!IS_ERR(dacl) && !IS_ERR(pacl)) {
+		set_cached_acl(inode, ACL_TYPE_DEFAULT, dacl);
+		set_cached_acl(inode, ACL_TYPE_ACCESS, pacl);
+		posix_acl_release(dacl);
+		posix_acl_release(pacl);
+	} else
+		retval = -EIO;
+
+	return retval;
+}
+
+static struct posix_acl *v9fs_get_cached_acl(struct inode *inode, int type)
+{
+	struct posix_acl *acl;
+	/*
+	 * 9p Always cache the acl value when
+	 * instantiating the inode (v9fs_inode_from_fid)
+	 */
+	acl = get_cached_acl(inode, type);
+	BUG_ON(acl == ACL_NOT_CACHED);
+	return acl;
+}
+
+int v9fs_check_acl(struct inode *inode, int mask)
+{
+	struct posix_acl *acl;
+	struct v9fs_session_info *v9ses;
+
+	v9ses = v9fs_inode2v9ses(inode);
+	if ((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT) {
+		/*
+		 * On access = client mode get the acl
+		 * values from the server
+		 */
+		return 0;
+	}
+	acl = v9fs_get_cached_acl(inode, ACL_TYPE_ACCESS);
+
+	if (IS_ERR(acl))
+		return PTR_ERR(acl);
+	if (acl) {
+		int error = posix_acl_permission(inode, acl, mask);
+		posix_acl_release(acl);
+		return error;
+	}
+	return -EAGAIN;
+}
+
+static int v9fs_set_acl(struct dentry *dentry, int type, struct posix_acl *acl)
+{
+	int retval;
+	char *name;
+	size_t size;
+	void *buffer;
+	struct inode *inode = dentry->d_inode;
+
+	set_cached_acl(inode, type, acl);
+	/* Set a setxattr request to server */
+	size = posix_acl_xattr_size(acl->a_count);
+	buffer = kmalloc(size, GFP_KERNEL);
+	if (!buffer)
+		return -ENOMEM;
+	retval = posix_acl_to_xattr(acl, buffer, size);
+	if (retval < 0)
+		goto err_free_out;
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		name = POSIX_ACL_XATTR_ACCESS;
+		break;
+	case ACL_TYPE_DEFAULT:
+		name = POSIX_ACL_XATTR_DEFAULT;
+		break;
+	default:
+		BUG();
+	}
+	retval = v9fs_xattr_set(dentry, name, buffer, size, 0);
+err_free_out:
+	kfree(buffer);
+	return retval;
+}
+
+int v9fs_acl_chmod(struct dentry *dentry)
+{
+	int retval = 0;
+	struct posix_acl *acl, *clone;
+	struct inode *inode = dentry->d_inode;
+
+	if (S_ISLNK(inode->i_mode))
+		return -EOPNOTSUPP;
+	acl = v9fs_get_cached_acl(inode, ACL_TYPE_ACCESS);
+	if (acl) {
+		clone = posix_acl_clone(acl, GFP_KERNEL);
+		posix_acl_release(acl);
+		if (!clone)
+			return -ENOMEM;
+		retval = posix_acl_chmod_masq(clone, inode->i_mode);
+		if (!retval)
+			retval = v9fs_set_acl(dentry, ACL_TYPE_ACCESS, clone);
+		posix_acl_release(clone);
+	}
+	return retval;
+}
+
+int v9fs_set_create_acl(struct dentry *dentry,
+			struct posix_acl *dpacl, struct posix_acl *pacl)
+{
+	if (dpacl)
+		v9fs_set_acl(dentry, ACL_TYPE_DEFAULT, dpacl);
+	if (pacl)
+		v9fs_set_acl(dentry, ACL_TYPE_ACCESS, pacl);
+	posix_acl_release(dpacl);
+	posix_acl_release(pacl);
+	return 0;
+}
+
+int v9fs_acl_mode(struct inode *dir, mode_t *modep,
+		  struct posix_acl **dpacl, struct posix_acl **pacl)
+{
+	int retval = 0;
+	mode_t mode = *modep;
+	struct posix_acl *acl = NULL;
+
+	if (!S_ISLNK(mode)) {
+		acl = v9fs_get_cached_acl(dir, ACL_TYPE_DEFAULT);
+		if (IS_ERR(acl))
+			return PTR_ERR(acl);
+		if (!acl)
+			mode &= ~current_umask();
+	}
+	if (acl) {
+		struct posix_acl *clone;
+
+		if (S_ISDIR(mode))
+			*dpacl = acl;
+		clone = posix_acl_clone(acl, GFP_NOFS);
+		retval = -ENOMEM;
+		if (!clone)
+			goto cleanup;
+
+		retval = posix_acl_create_masq(clone, &mode);
+		if (retval < 0) {
+			posix_acl_release(clone);
+			goto cleanup;
+		}
+		if (retval > 0)
+			*pacl = clone;
+	}
+	*modep  = mode;
+	return 0;
+cleanup:
+	posix_acl_release(acl);
+	return retval;
+
+}
+
+static int v9fs_remote_get_acl(struct dentry *dentry, const char *name,
+			       void *buffer, size_t size, int type)
+{
+	char *full_name;
+
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		full_name =  POSIX_ACL_XATTR_ACCESS;
+		break;
+	case ACL_TYPE_DEFAULT:
+		full_name = POSIX_ACL_XATTR_DEFAULT;
+		break;
+	default:
+		BUG();
+	}
+	return v9fs_xattr_get(dentry, full_name, buffer, size);
+}
+
+static int v9fs_xattr_get_acl(struct dentry *dentry, const char *name,
+			      void *buffer, size_t size, int type)
+{
+	struct v9fs_session_info *v9ses;
+	struct posix_acl *acl;
+	int error;
+
+	if (strcmp(name, "") != 0)
+		return -EINVAL;
+
+	v9ses = v9fs_inode2v9ses(dentry->d_inode);
+	/*
+	 * We allow set/get/list of acl when access=client is not specified
+	 */
+	if ((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT)
+		return v9fs_remote_get_acl(dentry, name, buffer, size, type);
+
+	acl = v9fs_get_cached_acl(dentry->d_inode, type);
+	if (IS_ERR(acl))
+		return PTR_ERR(acl);
+	if (acl == NULL)
+		return -ENODATA;
+	error = posix_acl_to_xattr(acl, buffer, size);
+	posix_acl_release(acl);
+
+	return error;
+}
+
+static int v9fs_remote_set_acl(struct dentry *dentry, const char *name,
+			      const void *value, size_t size,
+			      int flags, int type)
+{
+	char *full_name;
+
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		full_name =  POSIX_ACL_XATTR_ACCESS;
+		break;
+	case ACL_TYPE_DEFAULT:
+		full_name = POSIX_ACL_XATTR_DEFAULT;
+		break;
+	default:
+		BUG();
+	}
+	return v9fs_xattr_set(dentry, full_name, value, size, flags);
+}
+
+
+static int v9fs_xattr_set_acl(struct dentry *dentry, const char *name,
+			      const void *value, size_t size,
+			      int flags, int type)
+{
+	int retval;
+	struct posix_acl *acl;
+	struct v9fs_session_info *v9ses;
+	struct inode *inode = dentry->d_inode;
+
+	if (strcmp(name, "") != 0)
+		return -EINVAL;
+
+	v9ses = v9fs_inode2v9ses(dentry->d_inode);
+	/*
+	 * set the attribute on the remote. Without even looking at the
+	 * xattr value. We leave it to the server to validate
+	 */
+	if ((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT)
+		return v9fs_remote_set_acl(dentry, name,
+					   value, size, flags, type);
+
+	if (S_ISLNK(inode->i_mode))
+		return -EOPNOTSUPP;
+	if (!is_owner_or_cap(inode))
+		return -EPERM;
+	if (value) {
+		/* update the cached acl value */
+		acl = posix_acl_from_xattr(value, size);
+		if (IS_ERR(acl))
+			return PTR_ERR(acl);
+		else if (acl) {
+			retval = posix_acl_valid(acl);
+			if (retval)
+				goto err_out;
+		}
+	} else
+		acl = NULL;
+
+	switch (type) {
+	case ACL_TYPE_ACCESS:
+		name = POSIX_ACL_XATTR_ACCESS;
+		if (acl) {
+			mode_t mode = inode->i_mode;
+			retval = posix_acl_equiv_mode(acl, &mode);
+			if (retval < 0)
+				goto err_out;
+			else {
+				struct iattr iattr;
+				if (retval == 0) {
+					/*
+					 * ACL can be represented
+					 * by the mode bits. So don't
+					 * update ACL.
+					 */
+					acl = NULL;
+					value = NULL;
+					size = 0;
+				}
+				/* Updte the mode bits */
+				iattr.ia_mode = ((mode & S_IALLUGO) |
+						 (inode->i_mode & ~S_IALLUGO));
+				iattr.ia_valid = ATTR_MODE;
+				/* FIXME should we update ctime ?
+				 * What is the following setxattr update the
+				 * mode ?
+				 */
+				v9fs_vfs_setattr_dotl(dentry, &iattr);
+			}
+		}
+		break;
+	case ACL_TYPE_DEFAULT:
+		name = POSIX_ACL_XATTR_DEFAULT;
+		if (!S_ISDIR(inode->i_mode)) {
+			retval = -EINVAL;
+			goto err_out;
+		}
+		break;
+	default:
+		BUG();
+	}
+	retval = v9fs_xattr_set(dentry, name, value, size, flags);
+	if (!retval)
+		set_cached_acl(inode, type, acl);
+err_out:
+	posix_acl_release(acl);
+	return retval;
+}
+
+const struct xattr_handler v9fs_xattr_acl_access_handler = {
+	.prefix	= POSIX_ACL_XATTR_ACCESS,
+	.flags	= ACL_TYPE_ACCESS,
+	.get	= v9fs_xattr_get_acl,
+	.set	= v9fs_xattr_set_acl,
+};
+
+const struct xattr_handler v9fs_xattr_acl_default_handler = {
+	.prefix	= POSIX_ACL_XATTR_DEFAULT,
+	.flags	= ACL_TYPE_DEFAULT,
+	.get	= v9fs_xattr_get_acl,
+	.set	= v9fs_xattr_set_acl,
+};

fs/9p/acl.h

@@ -0,0 +1,49 @@
+/*
+ * Copyright IBM Corporation, 2010
+ * Author Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ */
+#ifndef FS_9P_ACL_H
+#define FS_9P_ACL_H
+
+#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);
+extern int v9fs_acl_chmod(struct dentry *);
+extern int v9fs_set_create_acl(struct dentry *,
+			       struct posix_acl *, struct posix_acl *);
+extern int v9fs_acl_mode(struct inode *dir, mode_t *modep,
+			 struct posix_acl **dpacl, struct posix_acl **pacl);
+#else
+#define v9fs_check_acl NULL
+static inline int v9fs_get_acl(struct inode *inode, struct p9_fid *fid)
+{
+	return 0;
+}
+static inline int v9fs_acl_chmod(struct dentry *dentry)
+{
+	return 0;
+}
+static inline int v9fs_set_create_acl(struct dentry *dentry,
+				      struct posix_acl *dpacl,
+				      struct posix_acl *pacl)
+{
+	return 0;
+}
+static inline int v9fs_acl_mode(struct inode *dir, mode_t *modep,
+				struct posix_acl **dpacl,
+				struct posix_acl **pacl)
+{
+	return 0;
+}
+
+#endif
+#endif /* FS_9P_XATTR_H */

fs/9p/fid.c

@@ -149,6 +149,7 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry)
 	switch (access) {
 	case V9FS_ACCESS_SINGLE:
 	case V9FS_ACCESS_USER:
+	case V9FS_ACCESS_CLIENT:
 		uid = current_fsuid();
 		any = 0;
 		break;

fs/9p/v9fs.c

@@ -193,7 +193,17 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts)
 				v9ses->flags |= V9FS_ACCESS_USER;
 			else if (strcmp(s, "any") == 0)
 				v9ses->flags |= V9FS_ACCESS_ANY;
-			else {
+			else if (strcmp(s, "client") == 0) {
+#ifdef CONFIG_9P_FS_POSIX_ACL
+				v9ses->flags |= V9FS_ACCESS_CLIENT;
+#else
+				P9_DPRINTK(P9_DEBUG_ERROR,
+					"access=client option not supported\n");
+				kfree(s);
+				ret = -EINVAL;
+				goto free_and_return;
+#endif
+			} else {
 				v9ses->flags |= V9FS_ACCESS_SINGLE;
 				v9ses->uid = simple_strtoul(s, &e, 10);
 				if (*e != '\0')
@@ -278,6 +288,16 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses,
 
 	v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ;
 
+	if (!v9fs_proto_dotl(v9ses) &&
+	    ((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_CLIENT)) {
+		/*
+		 * We support ACCESS_CLIENT only for dotl.
+		 * Fall back to ACCESS_USER
+		 */
+		v9ses->flags &= ~V9FS_ACCESS_MASK;
+		v9ses->flags |= V9FS_ACCESS_USER;
+	}
+	/*FIXME !! */
 	/* for legacy mode, fall back to V9FS_ACCESS_ANY */
 	if (!(v9fs_proto_dotu(v9ses) || v9fs_proto_dotl(v9ses)) &&
 		((v9ses->flags&V9FS_ACCESS_MASK) == V9FS_ACCESS_USER)) {

fs/9p/v9fs.h

@@ -33,13 +33,17 @@
  *
  * Session flags reflect options selected by users at mount time
  */
+#define	V9FS_ACCESS_ANY (V9FS_ACCESS_SINGLE | \
+			 V9FS_ACCESS_USER |   \
+			 V9FS_ACCESS_CLIENT)
+#define V9FS_ACCESS_MASK V9FS_ACCESS_ANY
+
 enum p9_session_flags {
 	V9FS_PROTO_2000U	= 0x01,
 	V9FS_PROTO_2000L	= 0x02,
 	V9FS_ACCESS_SINGLE	= 0x04,
 	V9FS_ACCESS_USER	= 0x08,
-	V9FS_ACCESS_ANY		= 0x0C,
+	V9FS_ACCESS_CLIENT	= 0x10
-	V9FS_ACCESS_MASK	= 0x0C,
 };
 
 /* possible values of ->cache */
@@ -113,8 +117,6 @@ void v9fs_session_close(struct v9fs_session_info *v9ses);
 void v9fs_session_cancel(struct v9fs_session_info *v9ses);
 void v9fs_session_begin_cancel(struct v9fs_session_info *v9ses);
 
-#define V9FS_MAGIC 0x01021997
-
 /* other default globals */
 #define V9FS_PORT	564
 #define V9FS_DEFUSER	"nobody"

fs/9p/v9fs_vfs.h

@@ -64,3 +64,7 @@ int v9fs_uflags2omode(int uflags, int extended);
 
 ssize_t v9fs_file_readn(struct file *, 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);
+
+#define P9_LOCK_TIMEOUT (30*HZ)

fs/9p/vfs_addr.c

@@ -154,10 +154,40 @@ static int v9fs_launder_page(struct page *page)
 	return 0;
 }
 
+/**
+ * v9fs_direct_IO - 9P address space operation for direct I/O
+ * @rw: direction (read or write)
+ * @iocb: target I/O control block
+ * @iov: array of vectors that define I/O buffer
+ * @pos: offset in file to begin the operation
+ * @nr_segs: size of iovec array
+ *
+ * The presence of v9fs_direct_IO() in the address space ops vector
+ * allowes open() O_DIRECT flags which would have failed otherwise.
+ *
+ * In the non-cached mode, we shunt off direct read and write requests before
+ * the VFS gets them, so this method should never be called.
+ *
+ * Direct IO is not 'yet' supported in the cached mode. Hence when
+ * this routine is called through generic_file_aio_read(), the read/write fails
+ * with an error.
+ *
+ */
+ssize_t v9fs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov,
+		loff_t pos, unsigned long nr_segs)
+{
+	P9_DPRINTK(P9_DEBUG_VFS, "v9fs_direct_IO: v9fs_direct_IO (%s) "
+			"off/no(%lld/%lu) EINVAL\n",
+			iocb->ki_filp->f_path.dentry->d_name.name,
+			(long long) pos, nr_segs);
+
+	return -EINVAL;
+}
 const struct address_space_operations v9fs_addr_operations = {
       .readpage = v9fs_vfs_readpage,
       .readpages = v9fs_vfs_readpages,
       .releasepage = v9fs_release_page,
       .invalidatepage = v9fs_invalidate_page,
       .launder_page = v9fs_launder_page,
+      .direct_IO = v9fs_direct_IO,
 };

fs/9p/vfs_dir.c

@@ -242,7 +242,8 @@ static int v9fs_dir_readdir_dotl(struct file *filp, void *dirent,
 		while (rdir->head < rdir->tail) {
 
 			err = p9dirent_read(rdir->buf + rdir->head,
-						buflen - rdir->head, &curdirent,
+						rdir->tail - rdir->head,
+						&curdirent,
 						fid->clnt->proto_version);
 			if (err < 0) {
 				P9_DPRINTK(P9_DEBUG_VFS, "returned %d\n", err);
@@ -314,4 +315,5 @@ const struct file_operations v9fs_dir_operations_dotl = {
 	.readdir = v9fs_dir_readdir_dotl,
 	.open = v9fs_file_open,
 	.release = v9fs_dir_release,
+        .fsync = v9fs_file_fsync_dotl,
 };

fs/9p/vfs_file.c

@@ -33,6 +33,7 @@
 #include <linux/inet.h>
 #include <linux/list.h>
 #include <linux/pagemap.h>
+#include <linux/utsname.h>
 #include <asm/uaccess.h>
 #include <linux/idr.h>
 #include <net/9p/9p.h>
@@ -44,6 +45,7 @@
 #include "cache.h"
 
 static const struct file_operations v9fs_cached_file_operations;
+static const struct file_operations v9fs_cached_file_operations_dotl;
 
 /**
  * v9fs_file_open - open a file (or directory)
@@ -92,6 +94,8 @@ int v9fs_file_open(struct inode *inode, struct file *file)
 		/* enable cached file options */
 		if(file->f_op == &v9fs_file_operations)
 			file->f_op = &v9fs_cached_file_operations;
+		else if (file->f_op == &v9fs_file_operations_dotl)
+			file->f_op = &v9fs_cached_file_operations_dotl;
 
 #ifdef CONFIG_9P_FSCACHE
 		v9fs_cache_inode_set_cookie(inode, file);
@@ -130,6 +134,206 @@ static int v9fs_file_lock(struct file *filp, int cmd, struct file_lock *fl)
 	return res;
 }
 
+static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
+{
+	struct p9_flock flock;
+	struct p9_fid *fid;
+	uint8_t status;
+	int res = 0;
+	unsigned char fl_type;
+
+	fid = filp->private_data;
+	BUG_ON(fid == NULL);
+
+	if ((fl->fl_flags & FL_POSIX) != FL_POSIX)
+		BUG();
+
+	res = posix_lock_file_wait(filp, fl);
+	if (res < 0)
+		goto out;
+
+	/* convert posix lock to p9 tlock args */
+	memset(&flock, 0, sizeof(flock));
+	flock.type = fl->fl_type;
+	flock.start = fl->fl_start;
+	if (fl->fl_end == OFFSET_MAX)
+		flock.length = 0;
+	else
+		flock.length = fl->fl_end - fl->fl_start + 1;
+	flock.proc_id = fl->fl_pid;
+	flock.client_id = utsname()->nodename;
+	if (IS_SETLKW(cmd))
+		flock.flags = P9_LOCK_FLAGS_BLOCK;
+
+	/*
+	 * if its a blocked request and we get P9_LOCK_BLOCKED as the status
+	 * for lock request, keep on trying
+	 */
+	for (;;) {
+		res = p9_client_lock_dotl(fid, &flock, &status);
+		if (res < 0)
+			break;
+
+		if (status != P9_LOCK_BLOCKED)
+			break;
+		if (status == P9_LOCK_BLOCKED && !IS_SETLKW(cmd))
+			break;
+		schedule_timeout_interruptible(P9_LOCK_TIMEOUT);
+	}
+
+	/* map 9p status to VFS status */
+	switch (status) {
+	case P9_LOCK_SUCCESS:
+		res = 0;
+		break;
+	case P9_LOCK_BLOCKED:
+		res = -EAGAIN;
+		break;
+	case P9_LOCK_ERROR:
+	case P9_LOCK_GRACE:
+		res = -ENOLCK;
+		break;
+	default:
+		BUG();
+	}
+
+	/*
+	 * incase server returned error for lock request, revert
+	 * it locally
+	 */
+	if (res < 0 && fl->fl_type != F_UNLCK) {
+		fl_type = fl->fl_type;
+		fl->fl_type = F_UNLCK;
+		res = posix_lock_file_wait(filp, fl);
+		fl->fl_type = fl_type;
+	}
+out:
+	return res;
+}
+
+static int v9fs_file_getlock(struct file *filp, struct file_lock *fl)
+{
+	struct p9_getlock glock;
+	struct p9_fid *fid;
+	int res = 0;
+
+	fid = filp->private_data;
+	BUG_ON(fid == NULL);
+
+	posix_test_lock(filp, fl);
+	/*
+	 * if we have a conflicting lock locally, no need to validate
+	 * with server
+	 */
+	if (fl->fl_type != F_UNLCK)
+		return res;
+
+	/* convert posix lock to p9 tgetlock args */
+	memset(&glock, 0, sizeof(glock));
+	glock.type = fl->fl_type;
+	glock.start = fl->fl_start;
+	if (fl->fl_end == OFFSET_MAX)
+		glock.length = 0;
+	else
+		glock.length = fl->fl_end - fl->fl_start + 1;
+	glock.proc_id = fl->fl_pid;
+	glock.client_id = utsname()->nodename;
+
+	res = p9_client_getlock_dotl(fid, &glock);
+	if (res < 0)
+		return res;
+	if (glock.type != F_UNLCK) {
+		fl->fl_type = glock.type;
+		fl->fl_start = glock.start;
+		if (glock.length == 0)
+			fl->fl_end = OFFSET_MAX;
+		else
+			fl->fl_end = glock.start + glock.length - 1;
+		fl->fl_pid = glock.proc_id;
+	} else
+		fl->fl_type = F_UNLCK;
+
+	return res;
+}
+
+/**
+ * v9fs_file_lock_dotl - lock a file (or directory)
+ * @filp: file to be locked
+ * @cmd: lock command
+ * @fl: file lock structure
+ *
+ */
+
+static int v9fs_file_lock_dotl(struct file *filp, int cmd, struct file_lock *fl)
+{
+	struct inode *inode = filp->f_path.dentry->d_inode;
+	int ret = -ENOLCK;
+
+	P9_DPRINTK(P9_DEBUG_VFS, "filp: %p cmd:%d lock: %p name: %s\n", filp,
+				cmd, fl, filp->f_path.dentry->d_name.name);
+
+	/* No mandatory locks */
+	if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
+		goto out_err;
+
+	if ((IS_SETLK(cmd) || IS_SETLKW(cmd)) && fl->fl_type != F_UNLCK) {
+		filemap_write_and_wait(inode->i_mapping);
+		invalidate_mapping_pages(&inode->i_data, 0, -1);
+	}
+
+	if (IS_SETLK(cmd) || IS_SETLKW(cmd))
+		ret = v9fs_file_do_lock(filp, cmd, fl);
+	else if (IS_GETLK(cmd))
+		ret = v9fs_file_getlock(filp, fl);
+	else
+		ret = -EINVAL;
+out_err:
+	return ret;
+}
+
+/**
+ * v9fs_file_flock_dotl - lock a file
+ * @filp: file to be locked
+ * @cmd: lock command
+ * @fl: file lock structure
+ *
+ */
+
+static int v9fs_file_flock_dotl(struct file *filp, int cmd,
+	struct file_lock *fl)
+{
+	struct inode *inode = filp->f_path.dentry->d_inode;
+	int ret = -ENOLCK;
+
+	P9_DPRINTK(P9_DEBUG_VFS, "filp: %p cmd:%d lock: %p name: %s\n", filp,
+				cmd, fl, filp->f_path.dentry->d_name.name);
+
+	/* No mandatory locks */
+	if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
+		goto out_err;
+
+	if (!(fl->fl_flags & FL_FLOCK))
+		goto out_err;
+
+	if ((IS_SETLK(cmd) || IS_SETLKW(cmd)) && fl->fl_type != F_UNLCK) {
+		filemap_write_and_wait(inode->i_mapping);
+		invalidate_mapping_pages(&inode->i_data, 0, -1);
+	}
+	/* Convert flock to posix lock */
+	fl->fl_owner = (fl_owner_t)filp;
+	fl->fl_start = 0;
+	fl->fl_end = OFFSET_MAX;
+	fl->fl_flags |= FL_POSIX;
+	fl->fl_flags ^= FL_FLOCK;
+
+	if (IS_SETLK(cmd) | IS_SETLKW(cmd))
+		ret = v9fs_file_do_lock(filp, cmd, fl);
+	else
+		ret = -EINVAL;
+out_err:
+	return ret;
+}
+
 /**
  * v9fs_file_readn - read from a file
  * @filp: file pointer to read
@@ -219,7 +423,9 @@ static ssize_t
 v9fs_file_write(struct file *filp, const char __user * data,
 		size_t count, loff_t * offset)
 {
-	int n, rsize, total = 0;
+	ssize_t retval;
+	size_t total = 0;
+	int n;
 	struct p9_fid *fid;
 	struct p9_client *clnt;
 	struct inode *inode = filp->f_path.dentry->d_inode;
@@ -232,14 +438,19 @@ v9fs_file_write(struct file *filp, const char __user * data,
 	fid = filp->private_data;
 	clnt = fid->clnt;
 
-	rsize = fid->iounit ? fid->iounit : clnt->msize - P9_IOHDRSZ;
+	retval = generic_write_checks(filp, &origin, &count, 0);
+	if (retval)
+		goto out;
+
+	retval = -EINVAL;
+	if ((ssize_t) count < 0)
+		goto out;
+	retval = 0;
+	if (!count)
+		goto out;
 
 	do {
-		if (count < rsize)
+		n = p9_client_write(fid, NULL, data+total, origin+total, count);
-			rsize = count;
-
-		n = p9_client_write(fid, NULL, data+total, origin+total,
-									rsize);
 		if (n <= 0)
 			break;
 		count -= n;
@@ -258,9 +469,11 @@ v9fs_file_write(struct file *filp, const char __user * data,
 	}
 
 	if (n < 0)
-		return n;
+		retval = n;
-
+	else
-	return total;
+		retval = total;
+out:
+	return retval;
 }
 
 static int v9fs_file_fsync(struct file *filp, int datasync)
@@ -278,6 +491,20 @@ static int v9fs_file_fsync(struct file *filp, int datasync)
 	return retval;
 }
 
+int v9fs_file_fsync_dotl(struct file *filp, int datasync)
+{
+	struct p9_fid *fid;
+	int retval;
+
+	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);
+	return retval;
+}
+
 static const struct file_operations v9fs_cached_file_operations = {
 	.llseek = generic_file_llseek,
 	.read = do_sync_read,
@@ -290,6 +517,19 @@ static const struct file_operations v9fs_cached_file_operations = {
 	.fsync = v9fs_file_fsync,
 };
 
+static const struct file_operations v9fs_cached_file_operations_dotl = {
+	.llseek = generic_file_llseek,
+	.read = do_sync_read,
+	.aio_read = generic_file_aio_read,
+	.write = v9fs_file_write,
+	.open = v9fs_file_open,
+	.release = v9fs_dir_release,
+	.lock = v9fs_file_lock_dotl,
+	.flock = v9fs_file_flock_dotl,
+	.mmap = generic_file_readonly_mmap,
+	.fsync = v9fs_file_fsync_dotl,
+};
+
 const struct file_operations v9fs_file_operations = {
 	.llseek = generic_file_llseek,
 	.read = v9fs_file_read,
@@ -307,7 +547,8 @@ const struct file_operations v9fs_file_operations_dotl = {
 	.write = v9fs_file_write,
 	.open = v9fs_file_open,
 	.release = v9fs_dir_release,
-	.lock = v9fs_file_lock,
+	.lock = v9fs_file_lock_dotl,
+	.flock = v9fs_file_flock_dotl,
 	.mmap = generic_file_readonly_mmap,
-	.fsync = v9fs_file_fsync,
+	.fsync = v9fs_file_fsync_dotl,
 };

fs/9p/vfs_inode.c

@@ -36,6 +36,7 @@
 #include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/xattr.h>
+#include <linux/posix_acl.h>
 #include <net/9p/9p.h>
 #include <net/9p/client.h>
 
@@ -44,6 +45,7 @@
 #include "fid.h"
 #include "cache.h"
 #include "xattr.h"
+#include "acl.h"
 
 static const struct inode_operations v9fs_dir_inode_operations;
 static const struct inode_operations v9fs_dir_inode_operations_dotu;
@@ -53,6 +55,10 @@ static const struct inode_operations v9fs_file_inode_operations_dotl;
 static const struct inode_operations v9fs_symlink_inode_operations;
 static const struct inode_operations v9fs_symlink_inode_operations_dotl;
 
+static int
+v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int omode,
+		    dev_t rdev);
+
 /**
  * unixmode2p9mode - convert unix mode bits to plan 9
  * @v9ses: v9fs session information
@@ -500,6 +506,11 @@ v9fs_inode_dotl(struct v9fs_session_info *v9ses, struct p9_fid *fid,
 	v9fs_vcookie_set_qid(ret, &st->qid);
 	v9fs_cache_inode_get_cookie(ret);
 #endif
+	err = v9fs_get_acl(ret, fid);
+	if (err) {
+		iput(ret);
+		goto error;
+	}
 	kfree(st);
 	return ret;
 error:
@@ -553,13 +564,6 @@ static int v9fs_remove(struct inode *dir, struct dentry *file, int rmdir)
 	return retval;
 }
 
-static int
-v9fs_open_created(struct inode *inode, struct file *file)
-{
-	return 0;
-}
-
-
 /**
  * v9fs_create - Create a file
  * @v9ses: session information
@@ -655,29 +659,37 @@ error:
  */
 
 static int
-v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int mode,
+v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode,
 		struct nameidata *nd)
 {
 	int err = 0;
 	char *name = NULL;
 	gid_t gid;
 	int flags;
+	mode_t mode;
 	struct v9fs_session_info *v9ses;
 	struct p9_fid *fid = NULL;
 	struct p9_fid *dfid, *ofid;
 	struct file *filp;
 	struct p9_qid qid;
 	struct inode *inode;
+	struct posix_acl *pacl = NULL, *dacl = NULL;
 
 	v9ses = v9fs_inode2v9ses(dir);
 	if (nd && nd->flags & LOOKUP_OPEN)
 		flags = nd->intent.open.flags - 1;
-	else
+	else {
-		flags = O_RDWR;
+		/*
+		 * create call without LOOKUP_OPEN is due
+		 * to mknod of regular files. So use mknod
+		 * operation.
+		 */
+		return v9fs_vfs_mknod_dotl(dir, dentry, omode, 0);
+	}
 
 	name = (char *) dentry->d_name.name;
 	P9_DPRINTK(P9_DEBUG_VFS, "v9fs_vfs_create_dotl: name:%s flags:0x%x "
-			"mode:0x%x\n", name, flags, mode);
+			"mode:0x%x\n", name, flags, omode);
 
 	dfid = v9fs_fid_lookup(dentry->d_parent);
 	if (IS_ERR(dfid)) {
@@ -695,6 +707,15 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int mode,
 	}
 
 	gid = v9fs_get_fsgid_for_create(dir);
+
+	mode = omode;
+	/* Update mode based on ACL value */
+	err = v9fs_acl_mode(dir, &mode, &dacl, &pacl);
+	if (err) {
+		P9_DPRINTK(P9_DEBUG_VFS,
+			   "Failed to get acl values in creat %d\n", err);
+		goto error;
+	}
 	err = p9_client_create_dotl(ofid, name, flags, mode, gid, &qid);
 	if (err < 0) {
 		P9_DPRINTK(P9_DEBUG_VFS,
@@ -702,46 +723,52 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int mode,
 				err);
 		goto error;
 	}
-
+	/* instantiate inode and assign the unopened fid to the dentry */
-	/* No need to populate the inode if we are not opening the file AND
+	if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE ||
-	 * not in cached mode.
+	    (nd && nd->flags & LOOKUP_OPEN)) {
-	 */
-	if (!v9ses->cache && !(nd && nd->flags & LOOKUP_OPEN)) {
-		/* Not in cached mode. No need to populate inode with stat */
-		dentry->d_op = &v9fs_dentry_operations;
-		p9_client_clunk(ofid);
-		d_instantiate(dentry, NULL);
-		return 0;
-	}
-
-	/* Now walk from the parent so we can get an unopened fid. */
 		fid = p9_client_walk(dfid, 1, &name, 1);
 		if (IS_ERR(fid)) {
 			err = PTR_ERR(fid);
-		P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n", err);
+			P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n",
+				err);
 			fid = NULL;
 			goto error;
 		}
 
-	/* instantiate inode and assign the unopened fid to dentry */
 		inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb);
 		if (IS_ERR(inode)) {
 			err = PTR_ERR(inode);
-		P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err);
+			P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n",
+				err);
 			goto error;
 		}
-	if (v9ses->cache)
 		dentry->d_op = &v9fs_cached_dentry_operations;
-	else
-		dentry->d_op = &v9fs_dentry_operations;
 		d_instantiate(dentry, inode);
 		err = v9fs_fid_add(dentry, fid);
 		if (err < 0)
 			goto error;
+		/* The fid would get clunked via a dput */
+		fid = NULL;
+	} else {
+		/*
+		 * Not in cached mode. No need to populate
+		 * inode with stat. We need to get an inode
+		 * so that we can set the acl with dentry
+		 */
+		inode = v9fs_get_inode(dir->i_sb, mode);
+		if (IS_ERR(inode)) {
+			err = PTR_ERR(inode);
+			goto error;
+		}
+		dentry->d_op = &v9fs_dentry_operations;
+		d_instantiate(dentry, inode);
+	}
+	/* Now set the ACL based on the default value */
+	v9fs_set_create_acl(dentry, dacl, pacl);
 
 	/* if we are opening a file, assign the open fid to the file */
 	if (nd && nd->flags & LOOKUP_OPEN) {
-		filp = lookup_instantiate_filp(nd, dentry, v9fs_open_created);
+		filp = lookup_instantiate_filp(nd, dentry, generic_file_open);
 		if (IS_ERR(filp)) {
 			p9_client_clunk(ofid);
 			return PTR_ERR(filp);
@@ -800,7 +827,7 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 
 	/* if we are opening a file, assign the open fid to the file */
 	if (nd && nd->flags & LOOKUP_OPEN) {
-		filp = lookup_instantiate_filp(nd, dentry, v9fs_open_created);
+		filp = lookup_instantiate_filp(nd, dentry, generic_file_open);
 		if (IS_ERR(filp)) {
 			err = PTR_ERR(filp);
 			goto error;
@@ -859,23 +886,28 @@ static int v9fs_vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
  *
  */
 
-static int v9fs_vfs_mkdir_dotl(struct inode *dir, struct dentry *dentry,
+static int v9fs_vfs_mkdir_dotl(struct inode *dir,
-					int mode)
+			       struct dentry *dentry, int omode)
 {
 	int err;
 	struct v9fs_session_info *v9ses;
 	struct p9_fid *fid = NULL, *dfid = NULL;
 	gid_t gid;
 	char *name;
+	mode_t mode;
 	struct inode *inode;
 	struct p9_qid qid;
 	struct dentry *dir_dentry;
+	struct posix_acl *dacl = NULL, *pacl = NULL;
 
 	P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", dentry->d_name.name);
 	err = 0;
 	v9ses = v9fs_inode2v9ses(dir);
 
-	mode |= S_IFDIR;
+	omode |= S_IFDIR;
+	if (dir->i_mode & S_ISGID)
+		omode |= S_ISGID;
+
 	dir_dentry = v9fs_dentry_from_dir_inode(dir);
 	dfid = v9fs_fid_lookup(dir_dentry);
 	if (IS_ERR(dfid)) {
@@ -886,11 +918,14 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, struct dentry *dentry,
 	}
 
 	gid = v9fs_get_fsgid_for_create(dir);
-	if (gid < 0) {
+	mode = omode;
-		P9_DPRINTK(P9_DEBUG_VFS, "v9fs_get_fsgid_for_create failed\n");
+	/* Update mode based on ACL value */
+	err = v9fs_acl_mode(dir, &mode, &dacl, &pacl);
+	if (err) {
+		P9_DPRINTK(P9_DEBUG_VFS,
+			   "Failed to get acl values in mkdir %d\n", err);
 		goto error;
 	}
-
 	name = (char *) dentry->d_name.name;
 	err = p9_client_mkdir_dotl(dfid, name, mode, gid, &qid);
 	if (err < 0)
@@ -920,7 +955,23 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, struct dentry *dentry,
 		if (err < 0)
 			goto error;
 		fid = NULL;
+	} else {
+		/*
+		 * Not in cached mode. No need to populate
+		 * inode with stat. We need to get an inode
+		 * so that we can set the acl with dentry
+		 */
+		inode = v9fs_get_inode(dir->i_sb, mode);
+		if (IS_ERR(inode)) {
+			err = PTR_ERR(inode);
+			goto error;
 		}
+		dentry->d_op = &v9fs_dentry_operations;
+		d_instantiate(dentry, inode);
+	}
+	/* Now set the ACL based on the default value */
+	v9fs_set_create_acl(dentry, dacl, pacl);
+
 error:
 	if (fid)
 		p9_client_clunk(fid);
@@ -979,7 +1030,7 @@ static struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry,
 
 	result = v9fs_fid_add(dentry, fid);
 	if (result < 0)
-		goto error;
+		goto error_iput;
 
 inst_out:
 	if (v9ses->cache)
@@ -990,6 +1041,8 @@ inst_out:
 	d_add(dentry, inode);
 	return NULL;
 
+error_iput:
+	iput(inode);
 error:
 	p9_client_clunk(fid);
 
@@ -1237,7 +1290,7 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr)
  *
  */
 
-static int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr)
+int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr)
 {
 	int retval;
 	struct v9fs_session_info *v9ses;
@@ -1279,6 +1332,12 @@ static int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr)
 
 	setattr_copy(dentry->d_inode, iattr);
 	mark_inode_dirty(dentry->d_inode);
+	if (iattr->ia_valid & ATTR_MODE) {
+		/* We also want to update ACL when we update mode bits */
+		retval = v9fs_acl_chmod(dentry);
+		if (retval < 0)
+			return retval;
+	}
 	return 0;
 }
 
@@ -1473,7 +1532,7 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen)
 	if (IS_ERR(fid))
 		return PTR_ERR(fid);
 
-	if (!v9fs_proto_dotu(v9ses) && !v9fs_proto_dotl(v9ses))
+	if (!v9fs_proto_dotu(v9ses))
 		return -EBADF;
 
 	st = p9_client_stat(fid);
@@ -1616,11 +1675,6 @@ v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry,
 
 	gid = v9fs_get_fsgid_for_create(dir);
 
-	if (gid < 0) {
-		P9_DPRINTK(P9_DEBUG_VFS, "v9fs_get_egid failed %d\n", gid);
-		goto error;
-	}
-
 	/* Server doesn't alter fid on TSYMLINK. Hence no need to clone it. */
 	err = p9_client_symlink(dfid, name, (char *)symname, gid, &qid);
 
@@ -1855,21 +1909,23 @@ v9fs_vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev)
  *
  */
 static int
-v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int mode,
+v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int omode,
 		dev_t rdev)
 {
 	int err;
 	char *name;
+	mode_t mode;
 	struct v9fs_session_info *v9ses;
 	struct p9_fid *fid = NULL, *dfid = NULL;
 	struct inode *inode;
 	gid_t gid;
 	struct p9_qid qid;
 	struct dentry *dir_dentry;
+	struct posix_acl *dacl = NULL, *pacl = NULL;
 
 	P9_DPRINTK(P9_DEBUG_VFS,
 		" %lu,%s mode: %x MAJOR: %u MINOR: %u\n", dir->i_ino,
-		dentry->d_name.name, mode, MAJOR(rdev), MINOR(rdev));
+		dentry->d_name.name, omode, MAJOR(rdev), MINOR(rdev));
 
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
@@ -1885,11 +1941,14 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int mode,
 	}
 
 	gid = v9fs_get_fsgid_for_create(dir);
-	if (gid < 0) {
+	mode = omode;
-		P9_DPRINTK(P9_DEBUG_VFS, "v9fs_get_fsgid_for_create failed\n");
+	/* Update mode based on ACL value */
+	err = v9fs_acl_mode(dir, &mode, &dacl, &pacl);
+	if (err) {
+		P9_DPRINTK(P9_DEBUG_VFS,
+			   "Failed to get acl values in mknod %d\n", err);
 		goto error;
 	}
-
 	name = (char *) dentry->d_name.name;
 
 	err = p9_client_mknod_dotl(dfid, name, mode, rdev, gid, &qid);
@@ -1933,13 +1992,68 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int mode,
 		dentry->d_op = &v9fs_dentry_operations;
 		d_instantiate(dentry, inode);
 	}
-
+	/* Now set the ACL based on the default value */
+	v9fs_set_create_acl(dentry, dacl, pacl);
 error:
 	if (fid)
 		p9_client_clunk(fid);
 	return err;
 }
 
+static int
+v9fs_vfs_readlink_dotl(struct dentry *dentry, char *buffer, int buflen)
+{
+	int retval;
+	struct p9_fid *fid;
+	char *target = NULL;
+
+	P9_DPRINTK(P9_DEBUG_VFS, " %s\n", dentry->d_name.name);
+	retval = -EPERM;
+	fid = v9fs_fid_lookup(dentry);
+	if (IS_ERR(fid))
+		return PTR_ERR(fid);
+
+	retval = p9_client_readlink(fid, &target);
+	if (retval < 0)
+		return retval;
+
+	strncpy(buffer, target, buflen);
+	P9_DPRINTK(P9_DEBUG_VFS, "%s -> %s\n", dentry->d_name.name, buffer);
+
+	retval = strnlen(buffer, buflen);
+	return retval;
+}
+
+/**
+ * v9fs_vfs_follow_link_dotl - follow a symlink path
+ * @dentry: dentry for symlink
+ * @nd: nameidata
+ *
+ */
+
+static void *
+v9fs_vfs_follow_link_dotl(struct dentry *dentry, struct nameidata *nd)
+{
+	int len = 0;
+	char *link = __getname();
+
+	P9_DPRINTK(P9_DEBUG_VFS, "%s n", dentry->d_name.name);
+
+	if (!link)
+		link = ERR_PTR(-ENOMEM);
+	else {
+		len = v9fs_vfs_readlink_dotl(dentry, link, PATH_MAX);
+		if (len < 0) {
+			__putname(link);
+			link = ERR_PTR(len);
+		} else
+			link[min(len, PATH_MAX-1)] = 0;
+	}
+	nd_set_link(nd, link);
+
+	return NULL;
+}
+
 static const struct inode_operations v9fs_dir_inode_operations_dotu = {
 	.create = v9fs_vfs_create,
 	.lookup = v9fs_vfs_lookup,
@@ -1970,7 +2084,7 @@ static const struct inode_operations v9fs_dir_inode_operations_dotl = {
 	.getxattr = generic_getxattr,
 	.removexattr = generic_removexattr,
 	.listxattr = v9fs_listxattr,
-
+	.check_acl = v9fs_check_acl,
 };
 
 static const struct inode_operations v9fs_dir_inode_operations = {
@@ -1997,6 +2111,7 @@ static const struct inode_operations v9fs_file_inode_operations_dotl = {
 	.getxattr = generic_getxattr,
 	.removexattr = generic_removexattr,
 	.listxattr = v9fs_listxattr,
+	.check_acl = v9fs_check_acl,
 };
 
 static const struct inode_operations v9fs_symlink_inode_operations = {
@@ -2008,8 +2123,8 @@ static const struct inode_operations v9fs_symlink_inode_operations = {
 };
 
 static const struct inode_operations v9fs_symlink_inode_operations_dotl = {
-	.readlink = generic_readlink,
+	.readlink = v9fs_vfs_readlink_dotl,
-	.follow_link = v9fs_vfs_follow_link,
+	.follow_link = v9fs_vfs_follow_link_dotl,
 	.put_link = v9fs_vfs_put_link,
 	.getattr = v9fs_vfs_getattr_dotl,
 	.setattr = v9fs_vfs_setattr_dotl,

fs/9p/vfs_super.c

@@ -39,6 +39,7 @@
 #include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/statfs.h>
+#include <linux/magic.h>
 #include <net/9p/9p.h>
 #include <net/9p/client.h>
 
@@ -46,6 +47,7 @@
 #include "v9fs_vfs.h"
 #include "fid.h"
 #include "xattr.h"
+#include "acl.h"
 
 static const struct super_operations v9fs_super_ops, v9fs_super_ops_dotl;
 
@@ -88,6 +90,11 @@ v9fs_fill_super(struct super_block *sb, struct v9fs_session_info *v9ses,
 	sb->s_flags = flags | MS_ACTIVE | MS_SYNCHRONOUS | MS_DIRSYNC |
 	    MS_NOATIME;
 
+#ifdef CONFIG_9P_FS_POSIX_ACL
+	if ((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_CLIENT)
+		sb->s_flags |= MS_POSIXACL;
+#endif
+
 	save_mount_options(sb, data);
 }
 
@@ -149,7 +156,6 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags,
 		goto release_sb;
 	}
 	sb->s_root = root;
-
 	if (v9fs_proto_dotl(v9ses)) {
 		struct p9_stat_dotl *st = NULL;
 		st = p9_client_getattr_dotl(fid, P9_STATS_BASIC);
@@ -174,7 +180,9 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags,
 		p9stat_free(st);
 		kfree(st);
 	}
-
+	retval = v9fs_get_acl(inode, fid);
+	if (retval)
+		goto release_sb;
 	v9fs_fid_add(root, fid);
 
 	P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n");
@@ -249,7 +257,7 @@ static int v9fs_statfs(struct dentry *dentry, struct kstatfs *buf)
 	if (v9fs_proto_dotl(v9ses)) {
 		res = p9_client_statfs(fid, &rs);
 		if (res == 0) {
-			buf->f_type = rs.type;
+			buf->f_type = V9FS_MAGIC;
 			buf->f_bsize = rs.bsize;
 			buf->f_blocks = rs.blocks;
 			buf->f_bfree = rs.bfree;

fs/9p/xattr.c

@@ -21,30 +21,13 @@
 #include "fid.h"
 #include "xattr.h"
 
-/*
+ssize_t v9fs_fid_xattr_get(struct p9_fid *fid, const char *name,
- * v9fs_xattr_get()
- *
- * Copy an extended attribute into the buffer
- * provided, or compute the buffer size required.
- * Buffer is NULL to compute the size of the buffer required.
- *
- * Returns a negative error number on failure, or the number of bytes
- * used / required on success.
- */
-ssize_t v9fs_xattr_get(struct dentry *dentry, const char *name,
 			   void *buffer, size_t buffer_size)
 {
 	ssize_t retval;
 	int msize, read_count;
 	u64 offset = 0, attr_size;
-	struct p9_fid *fid, *attr_fid;
+	struct p9_fid *attr_fid;
-
-	P9_DPRINTK(P9_DEBUG_VFS, "%s: name = %s value_len = %zu\n",
-		__func__, name, buffer_size);
-
-	fid = v9fs_fid_lookup(dentry);
-	if (IS_ERR(fid))
-		return PTR_ERR(fid);
 
 	attr_fid = p9_client_xattrwalk(fid, name, &attr_size);
 	if (IS_ERR(attr_fid)) {
@@ -88,6 +71,31 @@ error:
 
 }
 
+
+/*
+ * v9fs_xattr_get()
+ *
+ * Copy an extended attribute into the buffer
+ * provided, or compute the buffer size required.
+ * Buffer is NULL to compute the size of the buffer required.
+ *
+ * Returns a negative error number on failure, or the number of bytes
+ * used / required on success.
+ */
+ssize_t v9fs_xattr_get(struct dentry *dentry, const char *name,
+		       void *buffer, size_t buffer_size)
+{
+	struct p9_fid *fid;
+
+	P9_DPRINTK(P9_DEBUG_VFS, "%s: name = %s value_len = %zu\n",
+		__func__, name, buffer_size);
+	fid = v9fs_fid_lookup(dentry);
+	if (IS_ERR(fid))
+		return PTR_ERR(fid);
+
+	return v9fs_fid_xattr_get(fid, name, buffer, buffer_size);
+}
+
 /*
  * v9fs_xattr_set()
  *
@@ -156,5 +164,9 @@ ssize_t v9fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size)
 
 const struct xattr_handler *v9fs_xattr_handlers[] = {
 	&v9fs_xattr_user_handler,
+#ifdef CONFIG_9P_FS_POSIX_ACL
+	&v9fs_xattr_acl_access_handler,
+	&v9fs_xattr_acl_default_handler,
+#endif
 	NULL
 };

fs/9p/xattr.h

@@ -15,10 +15,16 @@
 #define FS_9P_XATTR_H
 
 #include <linux/xattr.h>
+#include <net/9p/9p.h>
+#include <net/9p/client.h>
 
 extern const struct xattr_handler *v9fs_xattr_handlers[];
 extern struct xattr_handler v9fs_xattr_user_handler;
+extern const struct xattr_handler v9fs_xattr_acl_access_handler;
+extern const struct xattr_handler v9fs_xattr_acl_default_handler;
 
+extern ssize_t v9fs_fid_xattr_get(struct p9_fid *, const char *,
+				  void *, size_t);
 extern ssize_t v9fs_xattr_get(struct dentry *, const char *,
 			      void *, size_t);
 extern int v9fs_xattr_set(struct dentry *, const char *,

include/linux/magic.h

@@ -57,5 +57,6 @@
 
 #define DEVPTS_SUPER_MAGIC	0x1cd1
 #define SOCKFS_MAGIC		0x534F434B
+#define V9FS_MAGIC		0x01021997
 
 #endif /* __LINUX_MAGIC_H__ */

include/net/9p/9p.h

@@ -86,6 +86,8 @@ do { \
 
 /**
  * enum p9_msg_t - 9P message types
+ * @P9_TLERROR: not used
+ * @P9_RLERROR: response for any failed request for 9P2000.L
  * @P9_TSTATFS: file system status request
  * @P9_RSTATFS: file system status response
  * @P9_TSYMLINK: make symlink request
@@ -137,6 +139,8 @@ do { \
  */
 
 enum p9_msg_t {
+	P9_TLERROR = 6,
+	P9_RLERROR,
 	P9_TSTATFS = 8,
 	P9_RSTATFS,
 	P9_TLOPEN = 12,
@@ -149,6 +153,8 @@ enum p9_msg_t {
 	P9_RMKNOD,
 	P9_TRENAME = 20,
 	P9_RRENAME,
+	P9_TREADLINK = 22,
+	P9_RREADLINK,
 	P9_TGETATTR = 24,
 	P9_RGETATTR,
 	P9_TSETATTR = 26,
@@ -159,6 +165,12 @@ enum p9_msg_t {
 	P9_RXATTRCREATE,
 	P9_TREADDIR = 40,
 	P9_RREADDIR,
+	P9_TFSYNC = 50,
+	P9_RFSYNC,
+	P9_TLOCK = 52,
+	P9_RLOCK,
+	P9_TGETLOCK = 54,
+	P9_RGETLOCK,
 	P9_TLINK = 70,
 	P9_RLINK,
 	P9_TMKDIR = 72,
@@ -458,6 +470,48 @@ struct p9_iattr_dotl {
 	u64 mtime_nsec;
 };
 
+#define P9_LOCK_SUCCESS 0
+#define P9_LOCK_BLOCKED 1
+#define P9_LOCK_ERROR 2
+#define P9_LOCK_GRACE 3
+
+#define P9_LOCK_FLAGS_BLOCK 1
+#define P9_LOCK_FLAGS_RECLAIM 2
+
+/* struct p9_flock: POSIX lock structure
+ * @type - type of lock
+ * @flags - lock flags
+ * @start - starting offset of the lock
+ * @length - number of bytes
+ * @proc_id - process id which wants to take lock
+ * @client_id - client id
+ */
+
+struct p9_flock {
+	u8 type;
+	u32 flags;
+	u64 start;
+	u64 length;
+	u32 proc_id;
+	char *client_id;
+};
+
+/* struct p9_getlock: getlock structure
+ * @type - type of lock
+ * @start - starting offset of the lock
+ * @length - number of bytes
+ * @proc_id - process id which wants to take lock
+ * @client_id - client id
+ */
+
+struct p9_getlock {
+	u8 type;
+	u64 start;
+	u64 length;
+	u32 proc_id;
+	char *client_id;
+};
+
 /* Structures for Protocol Operations */
 struct p9_tstatfs {
 	u32 fid;

include/net/9p/client.h

@@ -229,6 +229,7 @@ int p9_client_symlink(struct p9_fid *fid, char *name, char *symname, gid_t gid,
 int p9_client_create_dotl(struct p9_fid *ofid, char *name, u32 flags, u32 mode,
 		gid_t gid, struct p9_qid *qid);
 int p9_client_clunk(struct p9_fid *fid);
+int p9_client_fsync(struct p9_fid *fid, int datasync);
 int p9_client_remove(struct p9_fid *fid);
 int p9_client_read(struct p9_fid *fid, char *data, char __user *udata,
 							u64 offset, u32 count);
@@ -248,6 +249,8 @@ int p9_client_mknod_dotl(struct p9_fid *oldfid, char *name, int mode,
 			dev_t rdev, gid_t gid, struct p9_qid *);
 int p9_client_mkdir_dotl(struct p9_fid *fid, char *name, int mode,
 				gid_t gid, struct p9_qid *);
+int p9_client_lock_dotl(struct p9_fid *fid, struct p9_flock *flock, u8 *status);
+int p9_client_getlock_dotl(struct p9_fid *fid, struct p9_getlock *fl);
 struct p9_req_t *p9_tag_lookup(struct p9_client *, u16);
 void p9_client_cb(struct p9_client *c, struct p9_req_t *req);
 
@@ -259,5 +262,6 @@ int p9_is_proto_dotu(struct p9_client *clnt);
 int p9_is_proto_dotl(struct p9_client *clnt);
 struct p9_fid *p9_client_xattrwalk(struct p9_fid *, const char *, u64 *);
 int p9_client_xattrcreate(struct p9_fid *, const char *, u64, int);
+int p9_client_readlink(struct p9_fid *fid, char **target);
 
 #endif /* NET_9P_CLIENT_H */

net/9p/client.c

@@ -450,32 +450,43 @@ static int p9_check_errors(struct p9_client *c, struct p9_req_t *req)
 		return err;
 	}
 
-	if (type == P9_RERROR) {
+	if (type == P9_RERROR || type == P9_RLERROR) {
 		int ecode;
+
+		if (!p9_is_proto_dotl(c)) {
 			char *ename;
 
 			err = p9pdu_readf(req->rc, c->proto_version, "s?d",
 								&ename, &ecode);
-		if (err) {
+			if (err)
-			P9_DPRINTK(P9_DEBUG_ERROR, "couldn't parse error%d\n",
+				goto out_err;
-									err);
-			return err;
-		}
 
-		if (p9_is_proto_dotu(c) ||
+			if (p9_is_proto_dotu(c))
-			p9_is_proto_dotl(c))
 				err = -ecode;
 
-		if (!err || !IS_ERR_VALUE(err))
+			if (!err || !IS_ERR_VALUE(err)) {
 				err = p9_errstr2errno(ename, strlen(ename));
 
 				P9_DPRINTK(P9_DEBUG_9P, "<<< RERROR (%d) %s\n", -ecode, ename);
 
 				kfree(ename);
+			}
+		} else {
+			err = p9pdu_readf(req->rc, c->proto_version, "d", &ecode);
+			err = -ecode;
+
+			P9_DPRINTK(P9_DEBUG_9P, "<<< RLERROR (%d)\n", -ecode);
+		}
+
 	} else
 		err = 0;
 
 	return err;
+
+out_err:
+	P9_DPRINTK(P9_DEBUG_ERROR, "couldn't parse error%d\n", err);
+
+	return err;
 }
 
 /**
@@ -568,10 +579,13 @@ p9_client_rpc(struct p9_client *c, int8_t type, const char *fmt, ...)
 	va_start(ap, fmt);
 	err = p9pdu_vwritef(req->tc, c->proto_version, fmt, ap);
 	va_end(ap);
+	if (err)
+		goto reterr;
 	p9pdu_finalize(req->tc);
 
 	err = c->trans_mod->request(c, req);
 	if (err < 0) {
+		if (err != -ERESTARTSYS)
 			c->status = Disconnected;
 		goto reterr;
 	}
@@ -1151,12 +1165,44 @@ int p9_client_link(struct p9_fid *dfid, struct p9_fid *oldfid, char *newname)
 }
 EXPORT_SYMBOL(p9_client_link);
 
+int p9_client_fsync(struct p9_fid *fid, int datasync)
+{
+	int err;
+	struct p9_client *clnt;
+	struct p9_req_t *req;
+
+	P9_DPRINTK(P9_DEBUG_9P, ">>> TFSYNC fid %d datasync:%d\n",
+			fid->fid, datasync);
+	err = 0;
+	clnt = fid->clnt;
+
+	req = p9_client_rpc(clnt, P9_TFSYNC, "dd", fid->fid, datasync);
+	if (IS_ERR(req)) {
+		err = PTR_ERR(req);
+		goto error;
+	}
+
+	P9_DPRINTK(P9_DEBUG_9P, "<<< RFSYNC fid %d\n", fid->fid);
+
+	p9_free_req(clnt, req);
+
+error:
+	return err;
+}
+EXPORT_SYMBOL(p9_client_fsync);
+
 int p9_client_clunk(struct p9_fid *fid)
 {
 	int err;
 	struct p9_client *clnt;
 	struct p9_req_t *req;
 
+	if (!fid) {
+		P9_EPRINTK(KERN_WARNING, "Trying to clunk with NULL fid\n");
+		dump_stack();
+		return 0;
+	}
+
 	P9_DPRINTK(P9_DEBUG_9P, ">>> TCLUNK fid %d\n", fid->fid);
 	err = 0;
 	clnt = fid->clnt;
@@ -1240,16 +1286,13 @@ p9_client_read(struct p9_fid *fid, char *data, char __user *udata, u64 offset,
 
 	if (data) {
 		memmove(data, dataptr, count);
-	}
+	} else {
-
-	if (udata) {
 		err = copy_to_user(udata, dataptr, count);
 		if (err) {
 			err = -EFAULT;
 			goto free_and_error;
 		}
 	}
-
 	p9_free_req(clnt, req);
 	return count;
 
@@ -1761,3 +1804,96 @@ error:
 
 }
 EXPORT_SYMBOL(p9_client_mkdir_dotl);
+
+int p9_client_lock_dotl(struct p9_fid *fid, struct p9_flock *flock, u8 *status)
+{
+	int err;
+	struct p9_client *clnt;
+	struct p9_req_t *req;
+
+	err = 0;
+	clnt = fid->clnt;
+	P9_DPRINTK(P9_DEBUG_9P, ">>> TLOCK fid %d type %i flags %d "
+			"start %lld length %lld proc_id %d client_id %s\n",
+			fid->fid, flock->type, flock->flags, flock->start,
+			flock->length, flock->proc_id, flock->client_id);
+
+	req = p9_client_rpc(clnt, P9_TLOCK, "dbdqqds", fid->fid, flock->type,
+				flock->flags, flock->start, flock->length,
+					flock->proc_id, flock->client_id);
+
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+
+	err = p9pdu_readf(req->rc, clnt->proto_version, "b", status);
+	if (err) {
+		p9pdu_dump(1, req->rc);
+		goto error;
+	}
+	P9_DPRINTK(P9_DEBUG_9P, "<<< RLOCK status %i\n", *status);
+error:
+	p9_free_req(clnt, req);
+	return err;
+
+}
+EXPORT_SYMBOL(p9_client_lock_dotl);
+
+int p9_client_getlock_dotl(struct p9_fid *fid, struct p9_getlock *glock)
+{
+	int err;
+	struct p9_client *clnt;
+	struct p9_req_t *req;
+
+	err = 0;
+	clnt = fid->clnt;
+	P9_DPRINTK(P9_DEBUG_9P, ">>> TGETLOCK fid %d, type %i start %lld "
+		"length %lld proc_id %d client_id %s\n", fid->fid, glock->type,
+		glock->start, glock->length, glock->proc_id, glock->client_id);
+
+	req = p9_client_rpc(clnt, P9_TGETLOCK, "dbqqds", fid->fid,  glock->type,
+		glock->start, glock->length, glock->proc_id, glock->client_id);
+
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+
+	err = p9pdu_readf(req->rc, clnt->proto_version, "bqqds", &glock->type,
+			&glock->start, &glock->length, &glock->proc_id,
+			&glock->client_id);
+	if (err) {
+		p9pdu_dump(1, req->rc);
+		goto error;
+	}
+	P9_DPRINTK(P9_DEBUG_9P, "<<< RGETLOCK type %i start %lld length %lld "
+		"proc_id %d client_id %s\n", glock->type, glock->start,
+		glock->length, glock->proc_id, glock->client_id);
+error:
+	p9_free_req(clnt, req);
+	return err;
+}
+EXPORT_SYMBOL(p9_client_getlock_dotl);
+
+int p9_client_readlink(struct p9_fid *fid, char **target)
+{
+	int err;
+	struct p9_client *clnt;
+	struct p9_req_t *req;
+
+	err = 0;
+	clnt = fid->clnt;
+	P9_DPRINTK(P9_DEBUG_9P, ">>> TREADLINK fid %d\n", fid->fid);
+
+	req = p9_client_rpc(clnt, P9_TREADLINK, "d", fid->fid);
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+
+	err = p9pdu_readf(req->rc, clnt->proto_version, "s", target);
+	if (err) {
+		p9pdu_dump(1, req->rc);
+		goto error;
+	}
+	P9_DPRINTK(P9_DEBUG_9P, "<<< RREADLINK target %s\n", *target);
+error:
+	p9_free_req(clnt, req);
+	return err;
+}
+EXPORT_SYMBOL(p9_client_readlink);

net/9p/protocol.c

@@ -122,9 +122,8 @@ static size_t
 pdu_write_u(struct p9_fcall *pdu, const char __user *udata, size_t size)
 {
 	size_t len = MIN(pdu->capacity - pdu->size, size);
-	int err = copy_from_user(&pdu->sdata[pdu->size], udata, len);
+	if (copy_from_user(&pdu->sdata[pdu->size], udata, len))
-	if (err)
+		len = 0;
-		printk(KERN_WARNING "pdu_write_u returning: %d\n", err);
 
 	pdu->size += len;
 	return size - len;

net/9p/trans_virtio.c

@@ -75,6 +75,8 @@ struct virtio_chan {
 	struct p9_client *client;
 	struct virtio_device *vdev;
 	struct virtqueue *vq;
+	int ring_bufs_avail;
+	wait_queue_head_t *vc_wq;
 
 	/* Scatterlist: can be too big for stack. */
 	struct scatterlist sg[VIRTQUEUE_NUM];
@@ -134,16 +136,30 @@ static void req_done(struct virtqueue *vq)
 	struct p9_fcall *rc;
 	unsigned int len;
 	struct p9_req_t *req;
+	unsigned long flags;
 
 	P9_DPRINTK(P9_DEBUG_TRANS, ": request done\n");
 
-	while ((rc = virtqueue_get_buf(chan->vq, &len)) != NULL) {
+	do {
+		spin_lock_irqsave(&chan->lock, flags);
+		rc = virtqueue_get_buf(chan->vq, &len);
+
+		if (rc != NULL) {
+			if (!chan->ring_bufs_avail) {
+				chan->ring_bufs_avail = 1;
+				wake_up(chan->vc_wq);
+			}
+			spin_unlock_irqrestore(&chan->lock, flags);
 			P9_DPRINTK(P9_DEBUG_TRANS, ": rc %p\n", rc);
-		P9_DPRINTK(P9_DEBUG_TRANS, ": lookup tag %d\n", rc->tag);
+			P9_DPRINTK(P9_DEBUG_TRANS, ": lookup tag %d\n",
+					rc->tag);
 			req = p9_tag_lookup(chan->client, rc->tag);
 			req->status = REQ_STATUS_RCVD;
 			p9_client_cb(chan->client, req);
+		} else {
+			spin_unlock_irqrestore(&chan->lock, flags);
 		}
+	} while (rc != NULL);
 }
 
 /**
@@ -199,23 +215,43 @@ p9_virtio_request(struct p9_client *client, struct p9_req_t *req)
 	int in, out;
 	struct virtio_chan *chan = client->trans;
 	char *rdata = (char *)req->rc+sizeof(struct p9_fcall);
+	unsigned long flags;
+	int err;
 
 	P9_DPRINTK(P9_DEBUG_TRANS, "9p debug: virtio request\n");
 
+req_retry:
+	req->status = REQ_STATUS_SENT;
+
+	spin_lock_irqsave(&chan->lock, flags);
 	out = pack_sg_list(chan->sg, 0, VIRTQUEUE_NUM, req->tc->sdata,
 								req->tc->size);
 	in = pack_sg_list(chan->sg, out, VIRTQUEUE_NUM-out, rdata,
 								client->msize);
 
-	req->status = REQ_STATUS_SENT;
+	err = virtqueue_add_buf(chan->vq, chan->sg, out, in, req->tc);
+	if (err < 0) {
+		if (err == -ENOSPC) {
+			chan->ring_bufs_avail = 0;
+			spin_unlock_irqrestore(&chan->lock, flags);
+			err = wait_event_interruptible(*chan->vc_wq,
+							chan->ring_bufs_avail);
+			if (err  == -ERESTARTSYS)
+				return err;
 
-	if (virtqueue_add_buf(chan->vq, chan->sg, out, in, req->tc) < 0) {
+			P9_DPRINTK(P9_DEBUG_TRANS, "9p:Retry virtio request\n");
+			goto req_retry;
+		} else {
+			spin_unlock_irqrestore(&chan->lock, flags);
 			P9_DPRINTK(P9_DEBUG_TRANS,
-			"9p debug: virtio rpc add_buf returned failure");
+					"9p debug: "
+					"virtio rpc add_buf returned failure");
 			return -EIO;
 		}
+	}
 
 	virtqueue_kick(chan->vq);
+	spin_unlock_irqrestore(&chan->lock, flags);
 
 	P9_DPRINTK(P9_DEBUG_TRANS, "9p debug: virtio request kicked\n");
 	return 0;
@@ -290,14 +326,23 @@ static int p9_virtio_probe(struct virtio_device *vdev)
 	chan->tag_len = tag_len;
 	err = sysfs_create_file(&(vdev->dev.kobj), &dev_attr_mount_tag.attr);
 	if (err) {
-		kfree(tag);
+		goto out_free_tag;
-		goto out_free_vq;
 	}
+	chan->vc_wq = kmalloc(sizeof(wait_queue_head_t), GFP_KERNEL);
+	if (!chan->vc_wq) {
+		err = -ENOMEM;
+		goto out_free_tag;
+	}
+	init_waitqueue_head(chan->vc_wq);
+	chan->ring_bufs_avail = 1;
+
 	mutex_lock(&virtio_9p_lock);
 	list_add_tail(&chan->chan_list, &virtio_chan_list);
 	mutex_unlock(&virtio_9p_lock);
 	return 0;
 
+out_free_tag:
+	kfree(tag);
 out_free_vq:
 	vdev->config->del_vqs(vdev);
 	kfree(chan);
@@ -371,6 +416,7 @@ static void p9_virtio_remove(struct virtio_device *vdev)
 	mutex_unlock(&virtio_9p_lock);
 	sysfs_remove_file(&(vdev->dev.kobj), &dev_attr_mount_tag.attr);
 	kfree(chan->tag);
+	kfree(chan->vc_wq);
 	kfree(chan);
 
 }