2015-09-27 22:09:46

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 00/41] Richacls

Hello,

here's another update of the richacl patch queue. At this stage, I would
like to ask for final feedback so that the core and ext4 code (patches
1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
then go through the respective maintainer trees.

Changes since the last posting (https://lwn.net/Articles/656704/):

* The MAY_DELETE_SELF permission now also overrides the sticky
directory checks.

* Fix the permission check algorithm to apply the owner mask instead
of the group mask to user entries matching the current owner. That way,
the owner will retain the permissions in those entries when creating
objects with create mode 0700 and similar. (A chmod to mode 0700 already
creates an owner@:rwpx::allow ace, which was hiding this bug.)

* Fix richacl_apply_masks to properly insert deny aces when raising the
permissions of the other class. The bug could be triggered by
chmod'ing a group@:r::allow acl to mode 0077, for example.

* Various cleanups and improvements to comments.


The complete patch queue is available here:

git://git.kernel.org/pub/scm/linux/kernel/git/agruen/linux-richacl.git \
richacl-2015-09-28


The richacl user-space utilitites and test suite are available here:

https://github.com/andreas-gruenbacher/richacl/


Open issues in nfs:

* When a user or group name cannot be mapped, nfs's idmapper always maps it
to nobody. That's good enough for mapping the file owner and owning
group, but not for identifiers in acls. For now, to get the nfs richacl
support somewhat working, I'm explicitly checking if mapping has resulted
in uid/gid 99 in the kernel.

* When the nfs server replies with NFS4ERR_BADNAME for any user or group
name lookup, the client will stop sending numeric uids and gids to the
server even when the lookup wasn't numeric. From then on, the client
will translate uids and gids that have no mapping to the string "nobody",
and the server will reject them. This problem is not specific to acls.

Thanks,
Andreas

Andreas Gruenbacher (39):
vfs: Add IS_ACL() and IS_RICHACL() tests
vfs: Add MAY_CREATE_FILE and MAY_CREATE_DIR permission flags
vfs: Add MAY_DELETE_SELF and MAY_DELETE_CHILD permission flags
vfs: Make the inode passed to inode_change_ok non-const
vfs: Add permission flags for setting file attributes
richacl: In-memory representation and helper functions
richacl: Permission mapping functions
richacl: Compute maximum file masks from an acl
richacl: Update the file masks in chmod()
richacl: Permission check algorithm
vfs: Cache base_acl objects in inodes
vfs: Cache richacl in struct inode
richacl: Check if an acl is equivalent to a file mode
richacl: Create-time inheritance
richacl: Automatic Inheritance
richacl: xattr mapping functions
vfs: Add richacl permission checking
richacl: acl editing helper functions
richacl: Move everyone@ aces down the acl
richacl: Propagate everyone@ permissions to other aces
richacl: Set the owner permissions to the owner mask
richacl: Set the other permissions to the other mask
richacl: Isolate the owner and group classes
richacl: Apply the file masks to a richacl
richacl: Create richacl from mode values
nfsd: Keep list of acls to dispose of in compoundargs
nfsd: Use richacls as internal acl representation
nfsd: Add richacl support
nfsd: Add support for the v4.1 dacl attribute
nfsd: Add support for the MAY_CREATE_{FILE,DIR} permissions
richacl: Add support for unmapped identifiers
ext4: Don't allow unmapped identifiers in richacls
sunrpc: Allow to demand-allocate pages to encode into
sunrpc: Add xdr_init_encode_pages
nfs: Fix GETATTR bitmap verification
nfs: Remove unused xdr page offsets in getacl/setacl arguments
nfs: Add richacl support
nfs: Add support for the v4.1 dacl attribute
richacl: uapi header split

Aneesh Kumar K.V (2):
ext4: Add richacl support
ext4: Add richacl feature flag

drivers/staging/lustre/lustre/llite/llite_lib.c | 2 +-
fs/Kconfig | 9 +
fs/Makefile | 3 +
fs/attr.c | 81 ++-
fs/ext4/Kconfig | 15 +
fs/ext4/Makefile | 1 +
fs/ext4/acl.c | 6 +-
fs/ext4/acl.h | 12 +-
fs/ext4/ext4.h | 6 +-
fs/ext4/file.c | 6 +-
fs/ext4/ialloc.c | 7 +-
fs/ext4/inode.c | 10 +-
fs/ext4/namei.c | 11 +-
fs/ext4/richacl.c | 218 ++++++
fs/ext4/richacl.h | 47 ++
fs/ext4/super.c | 42 +-
fs/ext4/xattr.c | 6 +
fs/ext4/xattr.h | 1 +
fs/f2fs/acl.c | 4 +-
fs/inode.c | 15 +-
fs/jffs2/acl.c | 6 +-
fs/namei.c | 111 ++-
fs/nfs/inode.c | 3 -
fs/nfs/nfs4proc.c | 701 +++++++++++++-----
fs/nfs/nfs4xdr.c | 257 ++++++-
fs/nfs/super.c | 4 +-
fs/nfs_common/Makefile | 1 +
fs/nfs_common/nfs4acl.c | 44 ++
fs/nfsd/Kconfig | 1 +
fs/nfsd/acl.h | 23 +-
fs/nfsd/nfs4acl.c | 482 +++++++------
fs/nfsd/nfs4proc.c | 25 +-
fs/nfsd/nfs4xdr.c | 268 ++++---
fs/nfsd/nfsd.h | 6 +-
fs/nfsd/nfsfh.c | 8 +-
fs/nfsd/vfs.c | 28 +-
fs/nfsd/vfs.h | 17 +-
fs/nfsd/xdr4.h | 12 +-
fs/posix_acl.c | 26 +-
fs/richacl_base.c | 682 ++++++++++++++++++
fs/richacl_compat.c | 915 ++++++++++++++++++++++++
fs/richacl_inode.c | 297 ++++++++
fs/richacl_xattr.c | 267 +++++++
fs/xattr.c | 34 +-
include/linux/fs.h | 50 +-
include/linux/nfs4.h | 24 +-
include/linux/nfs4acl.h | 7 +
include/linux/nfs_fs.h | 1 -
include/linux/nfs_fs_sb.h | 2 +
include/linux/nfs_xdr.h | 13 +-
include/linux/posix_acl.h | 12 +-
include/linux/richacl.h | 275 +++++++
include/linux/richacl_compat.h | 40 ++
include/linux/richacl_xattr.h | 47 ++
include/linux/sunrpc/xdr.h | 2 +
include/uapi/linux/Kbuild | 2 +
include/uapi/linux/fs.h | 3 +-
include/uapi/linux/nfs4.h | 3 +-
include/uapi/linux/richacl.h | 111 +++
include/uapi/linux/richacl_xattr.h | 43 ++
include/uapi/linux/xattr.h | 2 +
net/sunrpc/xdr.c | 34 +
62 files changed, 4659 insertions(+), 732 deletions(-)
create mode 100644 fs/ext4/richacl.c
create mode 100644 fs/ext4/richacl.h
create mode 100644 fs/nfs_common/nfs4acl.c
create mode 100644 fs/richacl_base.c
create mode 100644 fs/richacl_compat.c
create mode 100644 fs/richacl_inode.c
create mode 100644 fs/richacl_xattr.c
create mode 100644 include/linux/nfs4acl.h
create mode 100644 include/linux/richacl.h
create mode 100644 include/linux/richacl_compat.h
create mode 100644 include/linux/richacl_xattr.h
create mode 100644 include/uapi/linux/richacl.h
create mode 100644 include/uapi/linux/richacl_xattr.h

--
2.4.3


Andreas Gruenbacher (39):
vfs: Add IS_ACL() and IS_RICHACL() tests
vfs: Add MAY_CREATE_FILE and MAY_CREATE_DIR permission flags
vfs: Add MAY_DELETE_SELF and MAY_DELETE_CHILD permission flags
vfs: Make the inode passed to inode_change_ok non-const
vfs: Add permission flags for setting file attributes
richacl: In-memory representation and helper functions
richacl: Permission mapping functions
richacl: Compute maximum file masks from an acl
richacl: Update the file masks in chmod()
richacl: Permission check algorithm
vfs: Cache base_acl objects in inodes
vfs: Cache richacl in struct inode
richacl: Check if an acl is equivalent to a file mode
richacl: Create-time inheritance
richacl: Automatic Inheritance
richacl: xattr mapping functions
vfs: Add richacl permission checking
richacl: acl editing helper functions
richacl: Move everyone@ aces down the acl
richacl: Propagate everyone@ permissions to other aces
richacl: Set the owner permissions to the owner mask
richacl: Set the other permissions to the other mask
richacl: Isolate the owner and group classes
richacl: Apply the file masks to a richacl
richacl: Create richacl from mode values
nfsd: Keep list of acls to dispose of in compoundargs
nfsd: Use richacls as internal acl representation
nfsd: Add richacl support
nfsd: Add support for the v4.1 dacl attribute
nfsd: Add support for the MAY_CREATE_{FILE,DIR} permissions
richacl: Add support for unmapped identifiers
ext4: Don't allow unmapped identifiers in richacls
sunrpc: Allow to demand-allocate pages to encode into
sunrpc: Add xdr_init_encode_pages
nfs: Fix GETATTR bitmap verification
nfs: Remove unused xdr page offsets in getacl/setacl arguments
nfs: Add richacl support
nfs: Add support for the v4.1 dacl attribute
richacl: uapi header split

Aneesh Kumar K.V (2):
ext4: Add richacl support
ext4: Add richacl feature flag

drivers/staging/lustre/lustre/llite/llite_lib.c | 2 +-
fs/Kconfig | 9 +
fs/Makefile | 3 +
fs/attr.c | 81 ++-
fs/ext4/Kconfig | 15 +
fs/ext4/Makefile | 1 +
fs/ext4/acl.c | 6 +-
fs/ext4/acl.h | 12 +-
fs/ext4/ext4.h | 6 +-
fs/ext4/file.c | 6 +-
fs/ext4/ialloc.c | 7 +-
fs/ext4/inode.c | 10 +-
fs/ext4/namei.c | 11 +-
fs/ext4/richacl.c | 218 ++++++
fs/ext4/richacl.h | 47 ++
fs/ext4/super.c | 42 +-
fs/ext4/xattr.c | 6 +
fs/ext4/xattr.h | 1 +
fs/f2fs/acl.c | 4 +-
fs/inode.c | 15 +-
fs/jffs2/acl.c | 6 +-
fs/namei.c | 111 ++-
fs/nfs/inode.c | 3 -
fs/nfs/nfs4proc.c | 701 +++++++++++++-----
fs/nfs/nfs4xdr.c | 257 ++++++-
fs/nfs/super.c | 4 +-
fs/nfs_common/Makefile | 1 +
fs/nfs_common/nfs4acl.c | 44 ++
fs/nfsd/Kconfig | 1 +
fs/nfsd/acl.h | 23 +-
fs/nfsd/nfs4acl.c | 482 +++++++------
fs/nfsd/nfs4proc.c | 25 +-
fs/nfsd/nfs4xdr.c | 268 ++++---
fs/nfsd/nfsd.h | 6 +-
fs/nfsd/nfsfh.c | 8 +-
fs/nfsd/vfs.c | 28 +-
fs/nfsd/vfs.h | 17 +-
fs/nfsd/xdr4.h | 12 +-
fs/posix_acl.c | 26 +-
fs/richacl_base.c | 682 ++++++++++++++++++
fs/richacl_compat.c | 915 ++++++++++++++++++++++++
fs/richacl_inode.c | 297 ++++++++
fs/richacl_xattr.c | 267 +++++++
fs/xattr.c | 34 +-
include/linux/fs.h | 50 +-
include/linux/nfs4.h | 24 +-
include/linux/nfs4acl.h | 7 +
include/linux/nfs_fs.h | 1 -
include/linux/nfs_fs_sb.h | 2 +
include/linux/nfs_xdr.h | 13 +-
include/linux/posix_acl.h | 12 +-
include/linux/richacl.h | 275 +++++++
include/linux/richacl_compat.h | 40 ++
include/linux/richacl_xattr.h | 47 ++
include/linux/sunrpc/xdr.h | 2 +
include/uapi/linux/Kbuild | 2 +
include/uapi/linux/fs.h | 3 +-
include/uapi/linux/nfs4.h | 3 +-
include/uapi/linux/richacl.h | 111 +++
include/uapi/linux/richacl_xattr.h | 43 ++
include/uapi/linux/xattr.h | 2 +
net/sunrpc/xdr.c | 34 +
62 files changed, 4659 insertions(+), 732 deletions(-)
create mode 100644 fs/ext4/richacl.c
create mode 100644 fs/ext4/richacl.h
create mode 100644 fs/nfs_common/nfs4acl.c
create mode 100644 fs/richacl_base.c
create mode 100644 fs/richacl_compat.c
create mode 100644 fs/richacl_inode.c
create mode 100644 fs/richacl_xattr.c
create mode 100644 include/linux/nfs4acl.h
create mode 100644 include/linux/richacl.h
create mode 100644 include/linux/richacl_compat.h
create mode 100644 include/linux/richacl_xattr.h
create mode 100644 include/uapi/linux/richacl.h
create mode 100644 include/uapi/linux/richacl_xattr.h

--
2.4.3



2015-09-27 22:13:40

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 41/41] richacl: uapi header split

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
include/linux/richacl.h | 119 +++++--------------------------------
include/linux/richacl_xattr.h | 17 +-----
include/uapi/linux/Kbuild | 2 +
include/uapi/linux/richacl.h | 111 ++++++++++++++++++++++++++++++++++
include/uapi/linux/richacl_xattr.h | 43 ++++++++++++++
5 files changed, 173 insertions(+), 119 deletions(-)
create mode 100644 include/uapi/linux/richacl.h
create mode 100644 include/uapi/linux/richacl_xattr.h

diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index 1822666..8875941 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -17,9 +17,7 @@
#ifndef __RICHACL_H
#define __RICHACL_H

-#define RICHACE_OWNER_SPECIAL_ID 0
-#define RICHACE_GROUP_SPECIAL_ID 1
-#define RICHACE_EVERYONE_SPECIAL_ID 2
+#include <uapi/linux/richacl.h>

struct richace {
unsigned short e_type;
@@ -44,43 +42,12 @@ struct richacl {
struct richace a_entries[0];
};

-#define richacl_for_each_entry(_ace, _acl) \
- for (_ace = (_acl)->a_entries; \
- _ace != (_acl)->a_entries + (_acl)->a_count; \
- _ace++)
-
-#define richacl_for_each_entry_reverse(_ace, _acl) \
- for (_ace = (_acl)->a_entries + (_acl)->a_count - 1; \
- _ace != (_acl)->a_entries - 1; \
- _ace--)
-
-/* a_flags values */
-#define RICHACL_AUTO_INHERIT 0x01
-#define RICHACL_PROTECTED 0x02
-#define RICHACL_DEFAULTED 0x04
-#define RICHACL_WRITE_THROUGH 0x40
-#define RICHACL_MASKED 0x80
-
#define RICHACL_VALID_FLAGS ( \
- RICHACL_AUTO_INHERIT | \
- RICHACL_PROTECTED | \
- RICHACL_DEFAULTED | \
- RICHACL_WRITE_THROUGH | \
- RICHACL_MASKED)
-
-/* e_type values */
-#define RICHACE_ACCESS_ALLOWED_ACE_TYPE 0x0000
-#define RICHACE_ACCESS_DENIED_ACE_TYPE 0x0001
-
-/* e_flags bitflags */
-#define RICHACE_FILE_INHERIT_ACE 0x0001
-#define RICHACE_DIRECTORY_INHERIT_ACE 0x0002
-#define RICHACE_NO_PROPAGATE_INHERIT_ACE 0x0004
-#define RICHACE_INHERIT_ONLY_ACE 0x0008
-#define RICHACE_IDENTIFIER_GROUP 0x0040
-#define RICHACE_INHERITED_ACE 0x0080
-#define RICHACE_UNMAPPED_WHO 0x2000
-#define RICHACE_SPECIAL_WHO 0x4000
+ RICHACL_AUTO_INHERIT | \
+ RICHACL_PROTECTED | \
+ RICHACL_DEFAULTED | \
+ RICHACL_WRITE_THROUGH | \
+ RICHACL_MASKED)

#define RICHACE_VALID_FLAGS ( \
RICHACE_FILE_INHERIT_ACE | \
@@ -99,27 +66,6 @@ struct richacl {
RICHACE_INHERIT_ONLY_ACE | \
RICHACE_INHERITED_ACE )

-/* e_mask bitflags */
-#define RICHACE_READ_DATA 0x00000001
-#define RICHACE_LIST_DIRECTORY 0x00000001
-#define RICHACE_WRITE_DATA 0x00000002
-#define RICHACE_ADD_FILE 0x00000002
-#define RICHACE_APPEND_DATA 0x00000004
-#define RICHACE_ADD_SUBDIRECTORY 0x00000004
-#define RICHACE_READ_NAMED_ATTRS 0x00000008
-#define RICHACE_WRITE_NAMED_ATTRS 0x00000010
-#define RICHACE_EXECUTE 0x00000020
-#define RICHACE_DELETE_CHILD 0x00000040
-#define RICHACE_READ_ATTRIBUTES 0x00000080
-#define RICHACE_WRITE_ATTRIBUTES 0x00000100
-#define RICHACE_WRITE_RETENTION 0x00000200
-#define RICHACE_WRITE_RETENTION_HOLD 0x00000400
-#define RICHACE_DELETE 0x00010000
-#define RICHACE_READ_ACL 0x00020000
-#define RICHACE_WRITE_ACL 0x00040000
-#define RICHACE_WRITE_OWNER 0x00080000
-#define RICHACE_SYNCHRONIZE 0x00100000
-
/* Valid RICHACE_* flags for directories and non-directories */
#define RICHACE_VALID_MASK ( \
RICHACE_READ_DATA | RICHACE_LIST_DIRECTORY | \
@@ -139,49 +85,16 @@ struct richacl {
RICHACE_WRITE_OWNER | \
RICHACE_SYNCHRONIZE)

-/*
- * The POSIX permissions are supersets of the following NFSv4 permissions:
- *
- * - MAY_READ maps to READ_DATA or LIST_DIRECTORY, depending on the type
- * of the file system object.
- *
- * - MAY_WRITE maps to WRITE_DATA or RICHACE_APPEND_DATA for files, and to
- * ADD_FILE, RICHACE_ADD_SUBDIRECTORY, or RICHACE_DELETE_CHILD for directories.
- *
- * - MAY_EXECUTE maps to RICHACE_EXECUTE.
- *
- * (Some of these NFSv4 permissions have the same bit values.)
- */
-#define RICHACE_POSIX_MODE_READ ( \
- RICHACE_READ_DATA | \
- RICHACE_LIST_DIRECTORY)
-#define RICHACE_POSIX_MODE_WRITE ( \
- RICHACE_WRITE_DATA | \
- RICHACE_ADD_FILE | \
- RICHACE_APPEND_DATA | \
- RICHACE_ADD_SUBDIRECTORY | \
- RICHACE_DELETE_CHILD)
-#define RICHACE_POSIX_MODE_EXEC RICHACE_EXECUTE
-#define RICHACE_POSIX_MODE_ALL ( \
- RICHACE_POSIX_MODE_READ | \
- RICHACE_POSIX_MODE_WRITE | \
- RICHACE_POSIX_MODE_EXEC)
-/*
- * These permissions are always allowed
- * no matter what the acl says.
- */
-#define RICHACE_POSIX_ALWAYS_ALLOWED ( \
- RICHACE_SYNCHRONIZE | \
- RICHACE_READ_ATTRIBUTES | \
- RICHACE_READ_ACL)
-/*
- * The owner is implicitly granted
- * these permissions under POSIX.
- */
-#define RICHACE_POSIX_OWNER_ALLOWED ( \
- RICHACE_WRITE_ATTRIBUTES | \
- RICHACE_WRITE_OWNER | \
- RICHACE_WRITE_ACL)
+#define richacl_for_each_entry(_ace, _acl) \
+ for (_ace = (_acl)->a_entries; \
+ _ace != (_acl)->a_entries + (_acl)->a_count; \
+ _ace++)
+
+#define richacl_for_each_entry_reverse(_ace, _acl) \
+ for (_ace = (_acl)->a_entries + (_acl)->a_count - 1; \
+ _ace != (_acl)->a_entries - 1; \
+ _ace--)
+
/**
* richacl_get - grab another reference to a richacl handle
*/
diff --git a/include/linux/richacl_xattr.h b/include/linux/richacl_xattr.h
index f84cc21..eff36a3 100644
--- a/include/linux/richacl_xattr.h
+++ b/include/linux/richacl_xattr.h
@@ -17,24 +17,9 @@
#ifndef __RICHACL_XATTR_H
#define __RICHACL_XATTR_H

+#include <uapi/linux/richacl_xattr.h>
#include <linux/richacl.h>

-struct richace_xattr {
- __le16 e_type;
- __le16 e_flags;
- __le32 e_mask;
- __le32 e_id;
-};
-
-struct richacl_xattr {
- unsigned char a_version;
- unsigned char a_flags;
- __le16 a_count;
- __le32 a_owner_mask;
- __le32 a_group_mask;
- __le32 a_other_mask;
-};
-
#define RICHACL_XATTR_VERSION 0
#define RICHACL_XATTR_MAX_COUNT \
((XATTR_SIZE_MAX - sizeof(struct richacl_xattr)) / \
diff --git a/include/uapi/linux/Kbuild b/include/uapi/linux/Kbuild
index f7b2db4..18ad070 100644
--- a/include/uapi/linux/Kbuild
+++ b/include/uapi/linux/Kbuild
@@ -348,6 +348,8 @@ header-y += reboot.h
header-y += reiserfs_fs.h
header-y += reiserfs_xattr.h
header-y += resource.h
+header-y += richacl.h
+header-y += richacl_xattr.h
header-y += rfkill.h
header-y += romfs_fs.h
header-y += rose.h
diff --git a/include/uapi/linux/richacl.h b/include/uapi/linux/richacl.h
new file mode 100644
index 0000000..6887f88
--- /dev/null
+++ b/include/uapi/linux/richacl.h
@@ -0,0 +1,111 @@
+/*
+ * Copyright (C) 2006, 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef __UAPI_RICHACL_H
+#define __UAPI_RICHACL_H
+
+/* a_flags values */
+#define RICHACL_AUTO_INHERIT 0x01
+#define RICHACL_PROTECTED 0x02
+#define RICHACL_DEFAULTED 0x04
+#define RICHACL_WRITE_THROUGH 0x40
+#define RICHACL_MASKED 0x80
+
+/* e_type values */
+#define RICHACE_ACCESS_ALLOWED_ACE_TYPE 0x0000
+#define RICHACE_ACCESS_DENIED_ACE_TYPE 0x0001
+
+/* e_flags bitflags */
+#define RICHACE_FILE_INHERIT_ACE 0x0001
+#define RICHACE_DIRECTORY_INHERIT_ACE 0x0002
+#define RICHACE_NO_PROPAGATE_INHERIT_ACE 0x0004
+#define RICHACE_INHERIT_ONLY_ACE 0x0008
+#define RICHACE_IDENTIFIER_GROUP 0x0040
+#define RICHACE_INHERITED_ACE 0x0080
+#define RICHACE_UNMAPPED_WHO 0x2000
+#define RICHACE_SPECIAL_WHO 0x4000
+
+/* e_mask bitflags */
+#define RICHACE_READ_DATA 0x00000001
+#define RICHACE_LIST_DIRECTORY 0x00000001
+#define RICHACE_WRITE_DATA 0x00000002
+#define RICHACE_ADD_FILE 0x00000002
+#define RICHACE_APPEND_DATA 0x00000004
+#define RICHACE_ADD_SUBDIRECTORY 0x00000004
+#define RICHACE_READ_NAMED_ATTRS 0x00000008
+#define RICHACE_WRITE_NAMED_ATTRS 0x00000010
+#define RICHACE_EXECUTE 0x00000020
+#define RICHACE_DELETE_CHILD 0x00000040
+#define RICHACE_READ_ATTRIBUTES 0x00000080
+#define RICHACE_WRITE_ATTRIBUTES 0x00000100
+#define RICHACE_WRITE_RETENTION 0x00000200
+#define RICHACE_WRITE_RETENTION_HOLD 0x00000400
+#define RICHACE_DELETE 0x00010000
+#define RICHACE_READ_ACL 0x00020000
+#define RICHACE_WRITE_ACL 0x00040000
+#define RICHACE_WRITE_OWNER 0x00080000
+#define RICHACE_SYNCHRONIZE 0x00100000
+
+/* e_id values */
+#define RICHACE_OWNER_SPECIAL_ID 0
+#define RICHACE_GROUP_SPECIAL_ID 1
+#define RICHACE_EVERYONE_SPECIAL_ID 2
+
+/*
+ * The POSIX permissions are supersets of the following richacl permissions:
+ *
+ * - MAY_READ maps to READ_DATA or LIST_DIRECTORY, depending on the type
+ * of the file system object.
+ *
+ * - MAY_WRITE maps to WRITE_DATA or RICHACE_APPEND_DATA for files, and to
+ * ADD_FILE, RICHACE_ADD_SUBDIRECTORY, or RICHACE_DELETE_CHILD for directories.
+ *
+ * - MAY_EXECUTE maps to RICHACE_EXECUTE.
+ *
+ * (Some of these richacl permissions have the same bit values.)
+ */
+#define RICHACE_POSIX_MODE_READ ( \
+ RICHACE_READ_DATA | \
+ RICHACE_LIST_DIRECTORY)
+#define RICHACE_POSIX_MODE_WRITE ( \
+ RICHACE_WRITE_DATA | \
+ RICHACE_ADD_FILE | \
+ RICHACE_APPEND_DATA | \
+ RICHACE_ADD_SUBDIRECTORY | \
+ RICHACE_DELETE_CHILD)
+#define RICHACE_POSIX_MODE_EXEC RICHACE_EXECUTE
+#define RICHACE_POSIX_MODE_ALL ( \
+ RICHACE_POSIX_MODE_READ | \
+ RICHACE_POSIX_MODE_WRITE | \
+ RICHACE_POSIX_MODE_EXEC)
+
+/*
+ * These permissions are always allowed no matter what the acl says.
+ */
+#define RICHACE_POSIX_ALWAYS_ALLOWED ( \
+ RICHACE_SYNCHRONIZE | \
+ RICHACE_READ_ATTRIBUTES | \
+ RICHACE_READ_ACL)
+
+/*
+ * The owner is implicitly granted these permissions under POSIX.
+ */
+#define RICHACE_POSIX_OWNER_ALLOWED ( \
+ RICHACE_WRITE_ATTRIBUTES | \
+ RICHACE_WRITE_OWNER | \
+ RICHACE_WRITE_ACL)
+
+#endif /* __UAPI_RICHACL_H */
diff --git a/include/uapi/linux/richacl_xattr.h b/include/uapi/linux/richacl_xattr.h
new file mode 100644
index 0000000..6f96bc0
--- /dev/null
+++ b/include/uapi/linux/richacl_xattr.h
@@ -0,0 +1,43 @@
+/*
+ * Copyright (C) 2006, 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef __UAPI_RICHACL_XATTR_H
+#define __UAPI_RICHACL_XATTR_H
+
+#include <linux/types.h>
+#include <linux/xattr.h>
+
+struct richace_xattr {
+ __le16 e_type;
+ __le16 e_flags;
+ __le32 e_mask;
+ __le32 e_id;
+};
+
+struct richacl_xattr {
+ unsigned char a_version;
+ unsigned char a_flags;
+ __le16 a_count;
+ __le32 a_owner_mask;
+ __le32 a_group_mask;
+ __le32 a_other_mask;
+};
+
+#define RICHACL_XATTR_MAX_COUNT \
+ ((XATTR_SIZE_MAX - sizeof(struct richacl_xattr)) / \
+ sizeof(struct richace_xattr))
+
+#endif /* __UAPI_RICHACL_XATTR_H */
--
2.4.3


2015-09-27 22:13:35

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 40/41] nfs: Add support for the v4.1 dacl attribute

The dacl attribute includes Automatic Inheritance flags not supported by
the acl attribute. it is only supported in NFS version 4.1 and higher.
On systems where NFS version 4.0 is still the default, an additional
mount option is needed:

mount -t nfs4 -o vers=4.1 [...]

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/nfs/nfs4proc.c | 2 +-
fs/nfs/nfs4xdr.c | 55 ++++++++++++++++++++++++++++++++++++++++++-------
include/linux/nfs_xdr.h | 2 +-
3 files changed, 50 insertions(+), 9 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 3888c70..3cd2402 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -4545,7 +4545,7 @@ static struct richacl *__nfs4_get_acl_uncached(struct inode *inode)
struct nfs_server *server = NFS_SERVER(inode);
struct page *pages[DIV_ROUND_UP(NFS4ACL_SIZE_MAX, PAGE_SIZE)] = {};
struct nfs_getaclargs args = {
- .fh = NFS_FH(inode),
+ .inode = inode,
.acl_pages = pages,
.acl_len = ARRAY_SIZE(pages) * PAGE_SIZE,
};
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index f2507d7..d855c6b 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -1661,9 +1661,16 @@ encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg, struct compoun
encode_nfs4_stateid(xdr, &zero_stateid);

/* Encode attribute bitmap. */
- p = reserve_space(xdr, 2*4);
- *p++ = cpu_to_be32(1);
- *p = cpu_to_be32(FATTR4_WORD0_ACL);
+ if (arg->server->attr_bitmask[1] & FATTR4_WORD1_DACL) {
+ p = reserve_space(xdr, 3*4);
+ *p++ = cpu_to_be32(2);
+ *p++ = 0;
+ *p = cpu_to_be32(FATTR4_WORD1_DACL);
+ } else {
+ p = reserve_space(xdr, 2*4);
+ *p++ = cpu_to_be32(1);
+ *p = cpu_to_be32(FATTR4_WORD0_ACL);
+ }

attrlen_offset = xdr->buf->len;
xdr_reserve_space(xdr, 4); /* to be backfilled later */
@@ -2498,9 +2505,12 @@ static void nfs4_xdr_enc_getacl(struct rpc_rqst *req, struct xdr_stream *xdr,

encode_compound_hdr(xdr, req, &hdr);
encode_sequence(xdr, &args->seq_args, &hdr);
- encode_putfh(xdr, args->fh, &hdr);
+ encode_putfh(xdr, NFS_FH(args->inode), &hdr);
replen = hdr.replen + op_decode_hdr_maxsz + 1;
- encode_getattr_two(xdr, FATTR4_WORD0_ACL, FATTR4_WORD1_MODE, &hdr);
+ if (NFS_SERVER(args->inode)->attr_bitmask[1] & FATTR4_WORD1_DACL)
+ encode_getattr_two(xdr, 0, FATTR4_WORD1_MODE | FATTR4_WORD1_DACL, &hdr);
+ else
+ encode_getattr_two(xdr, FATTR4_WORD0_ACL, FATTR4_WORD1_MODE, &hdr);

xdr_inline_pages(&req->rq_rcv_buf, replen << 2,
args->acl_pages, 0, args->acl_len);
@@ -5408,14 +5418,28 @@ static int decode_getacl(struct xdr_stream *xdr, struct rpc_rqst *req,

if (unlikely(bitmap[0] & (FATTR4_WORD0_ACL - 1U)))
return -EIO;
- if (likely(bitmap[0] & FATTR4_WORD0_ACL)) {
+
+ if (bitmap[0] & FATTR4_WORD0_ACL) {
+ struct richace *ace;
+
+ if (bitmap[1] & FATTR4_WORD1_DACL)
+ return -EIO;
+
acl = decode_acl_entries(xdr, res->server);
status = PTR_ERR(acl);
if (IS_ERR(acl))
goto out;
+
+ status = -EIO;
+ richacl_for_each_entry(ace, acl) {
+ if (ace->e_flags & RICHACE_INHERITED_ACE)
+ goto out;
+ }
bitmap[0] &= ~FATTR4_WORD0_ACL;
- } else
+ } else if (!(bitmap[1] & FATTR4_WORD1_DACL)) {
status = -EOPNOTSUPP;
+ goto out;
+ }

status = -EIO;
if (unlikely(bitmap[0]))
@@ -5424,6 +5448,23 @@ static int decode_getacl(struct xdr_stream *xdr, struct rpc_rqst *req,
status = decode_attr_mode(xdr, bitmap, &res->mode);
if (status < 0)
goto out;
+ if (bitmap[1] & FATTR4_WORD1_DACL) {
+ unsigned int flags;
+ __be32 *p;
+
+ p = xdr_inline_decode(xdr, 4);
+ status = -EIO;
+ if (unlikely(!p))
+ goto out;
+ flags = be32_to_cpup(p);
+
+ acl = decode_acl_entries(xdr, res->server);
+ status = PTR_ERR(acl);
+ if (IS_ERR(acl))
+ goto out;
+ acl->a_flags = flags;
+ bitmap[1] &= ~FATTR4_WORD1_DACL;
+ }
status = 0;

out:
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 337c341..9c2a078 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -695,7 +695,7 @@ struct nfs_setaclres {

struct nfs_getaclargs {
struct nfs4_sequence_args seq_args;
- struct nfs_fh * fh;
+ struct inode * inode;
size_t acl_len;
struct page ** acl_pages;
};
--
2.4.3


2015-09-27 22:13:29

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 39/41] nfs: Add richacl support

Add support for the "system.richacl" xattr in nfs. The existing
"system.nfs4_acl" xattr on nfs doesn't map user and group names to uids
and gids; the "system.richacl" xattr does, and only keeps the
on-the-wire names when there is no mapping. This allows to copy
permissions across different file systems.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/nfs/inode.c | 3 -
fs/nfs/nfs4proc.c | 698 +++++++++++++++++++++++++++++++++-------------
fs/nfs/nfs4xdr.c | 179 ++++++++++--
fs/nfs/super.c | 4 +-
include/linux/nfs_fs.h | 1 -
include/linux/nfs_fs_sb.h | 2 +
include/linux/nfs_xdr.h | 9 +-
7 files changed, 673 insertions(+), 223 deletions(-)

diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 326d9e1..843d15d 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1852,9 +1852,6 @@ struct inode *nfs_alloc_inode(struct super_block *sb)
return NULL;
nfsi->flags = 0UL;
nfsi->cache_validity = 0UL;
-#if IS_ENABLED(CONFIG_NFS_V4)
- nfsi->nfs4_acl = NULL;
-#endif /* CONFIG_NFS_V4 */
return &nfsi->vfs_inode;
}
EXPORT_SYMBOL_GPL(nfs_alloc_inode);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b3a6558..3888c70 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -55,6 +55,9 @@
#include <linux/xattr.h>
#include <linux/utsname.h>
#include <linux/freezer.h>
+#include <linux/richacl.h>
+#include <linux/richacl_xattr.h>
+#include <linux/nfs4acl.h>

#include "nfs4_fs.h"
#include "delegation.h"
@@ -2971,15 +2974,18 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f
res.attr_bitmask[2] &= FATTR4_WORD2_NFS42_MASK;
}
memcpy(server->attr_bitmask, res.attr_bitmask, sizeof(server->attr_bitmask));
- server->caps &= ~(NFS_CAP_ACLS|NFS_CAP_HARDLINKS|
- NFS_CAP_SYMLINKS|NFS_CAP_FILEID|
+ server->caps &= ~(NFS_CAP_ALLOW_ACLS|NFS_CAP_DENY_ACLS|
+ NFS_CAP_HARDLINKS|NFS_CAP_SYMLINKS|NFS_CAP_FILEID|
NFS_CAP_MODE|NFS_CAP_NLINK|NFS_CAP_OWNER|
NFS_CAP_OWNER_GROUP|NFS_CAP_ATIME|
NFS_CAP_CTIME|NFS_CAP_MTIME|
NFS_CAP_SECURITY_LABEL);
- if (res.attr_bitmask[0] & FATTR4_WORD0_ACL &&
- res.acl_bitmask & ACL4_SUPPORT_ALLOW_ACL)
- server->caps |= NFS_CAP_ACLS;
+ if (res.attr_bitmask[0] & FATTR4_WORD0_ACL) {
+ if (res.acl_bitmask & ACL4_SUPPORT_ALLOW_ACL)
+ server->caps |= NFS_CAP_ALLOW_ACLS;
+ if (res.acl_bitmask & ACL4_SUPPORT_DENY_ACL)
+ server->caps |= NFS_CAP_DENY_ACLS;
+ }
if (res.has_links != 0)
server->caps |= NFS_CAP_HARDLINKS;
if (res.has_symlinks != 0)
@@ -4507,45 +4513,11 @@ static int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred)
return 0;
}

-static inline int nfs4_server_supports_acls(struct nfs_server *server)
-{
- return server->caps & NFS_CAP_ACLS;
-}
-
-/* Assuming that XATTR_SIZE_MAX is a multiple of PAGE_SIZE, and that
- * it's OK to put sizeof(void) * (XATTR_SIZE_MAX/PAGE_SIZE) bytes on
- * the stack.
+/* A arbitrary limit; we allocate at most DIV_ROUND_UP(NFS4ACL_SIZE_MAX,
+ * PAGE_SIZE) pages and put an array of DIV_ROUND_UP(NFS4ACL_SIZE_MAX,
+ * PAGE_SIZE) pages on the stack when encoding or decoding acls.
*/
-#define NFS4ACL_MAXPAGES DIV_ROUND_UP(XATTR_SIZE_MAX, PAGE_SIZE)
-
-static int buf_to_pages_noslab(const void *buf, size_t buflen,
- struct page **pages)
-{
- struct page *newpage, **spages;
- int rc = 0;
- size_t len;
- spages = pages;
-
- do {
- len = min_t(size_t, PAGE_SIZE, buflen);
- newpage = alloc_page(GFP_KERNEL);
-
- if (newpage == NULL)
- goto unwind;
- memcpy(page_address(newpage), buf, len);
- buf += len;
- buflen -= len;
- *pages++ = newpage;
- rc++;
- } while (buflen != 0);
-
- return rc;
-
-unwind:
- for(; rc > 0; rc--)
- __free_page(spages[rc-1]);
- return -ENOMEM;
-}
+#define NFS4ACL_SIZE_MAX 65536

struct nfs4_cached_acl {
int cached;
@@ -4553,66 +4525,9 @@ struct nfs4_cached_acl {
char data[0];
};

-static void nfs4_set_cached_acl(struct inode *inode, struct nfs4_cached_acl *acl)
-{
- struct nfs_inode *nfsi = NFS_I(inode);
-
- spin_lock(&inode->i_lock);
- kfree(nfsi->nfs4_acl);
- nfsi->nfs4_acl = acl;
- spin_unlock(&inode->i_lock);
-}
-
static void nfs4_zap_acl_attr(struct inode *inode)
{
- nfs4_set_cached_acl(inode, NULL);
-}
-
-static inline ssize_t nfs4_read_cached_acl(struct inode *inode, char *buf, size_t buflen)
-{
- struct nfs_inode *nfsi = NFS_I(inode);
- struct nfs4_cached_acl *acl;
- int ret = -ENOENT;
-
- spin_lock(&inode->i_lock);
- acl = nfsi->nfs4_acl;
- if (acl == NULL)
- goto out;
- if (buf == NULL) /* user is just asking for length */
- goto out_len;
- if (acl->cached == 0)
- goto out;
- ret = -ERANGE; /* see getxattr(2) man page */
- if (acl->len > buflen)
- goto out;
- memcpy(buf, acl->data, acl->len);
-out_len:
- ret = acl->len;
-out:
- spin_unlock(&inode->i_lock);
- return ret;
-}
-
-static void nfs4_write_cached_acl(struct inode *inode, struct page **pages, size_t pgbase, size_t acl_len)
-{
- struct nfs4_cached_acl *acl;
- size_t buflen = sizeof(*acl) + acl_len;
-
- if (buflen <= PAGE_SIZE) {
- acl = kmalloc(buflen, GFP_KERNEL);
- if (acl == NULL)
- goto out;
- acl->cached = 1;
- _copy_from_pages(acl->data, pages, pgbase, acl_len);
- } else {
- acl = kmalloc(sizeof(*acl), GFP_KERNEL);
- if (acl == NULL)
- goto out;
- acl->cached = 0;
- }
- acl->len = acl_len;
-out:
- nfs4_set_cached_acl(inode, acl);
+ forget_cached_richacl(inode);
}

/*
@@ -4625,121 +4540,269 @@ out:
* length. The next getxattr call will then produce another round trip to
* the server, this time with the input buf of the required size.
*/
-static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t buflen)
+static struct richacl *__nfs4_get_acl_uncached(struct inode *inode)
{
- struct page *pages[NFS4ACL_MAXPAGES] = {NULL, };
+ struct nfs_server *server = NFS_SERVER(inode);
+ struct page *pages[DIV_ROUND_UP(NFS4ACL_SIZE_MAX, PAGE_SIZE)] = {};
struct nfs_getaclargs args = {
.fh = NFS_FH(inode),
.acl_pages = pages,
- .acl_len = buflen,
+ .acl_len = ARRAY_SIZE(pages) * PAGE_SIZE,
};
struct nfs_getaclres res = {
- .acl_len = buflen,
+ .server = server,
};
struct rpc_message msg = {
.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_GETACL],
.rpc_argp = &args,
.rpc_resp = &res,
};
- unsigned int npages = DIV_ROUND_UP(buflen, PAGE_SIZE);
- int ret = -ENOMEM, i;
+ int err, i;

- /* As long as we're doing a round trip to the server anyway,
- * let's be prepared for a page of acl data. */
- if (npages == 0)
- npages = 1;
- if (npages > ARRAY_SIZE(pages))
- return -ERANGE;
-
- for (i = 0; i < npages; i++) {
- pages[i] = alloc_page(GFP_KERNEL);
- if (!pages[i])
+ if (ARRAY_SIZE(pages) > 1) {
+ /* for decoding across pages */
+ res.acl_scratch = alloc_page(GFP_KERNEL);
+ err = -ENOMEM;
+ if (!res.acl_scratch)
goto out_free;
}

- /* for decoding across pages */
- res.acl_scratch = alloc_page(GFP_KERNEL);
- if (!res.acl_scratch)
- goto out_free;
-
- args.acl_len = npages * PAGE_SIZE;
-
- dprintk("%s buf %p buflen %zu npages %d args.acl_len %zu\n",
- __func__, buf, buflen, npages, args.acl_len);
- ret = nfs4_call_sync(NFS_SERVER(inode)->client, NFS_SERVER(inode),
+ dprintk("%s args.acl_len %zu\n",
+ __func__, args.acl_len);
+ err = nfs4_call_sync(NFS_SERVER(inode)->client, NFS_SERVER(inode),
&msg, &args.seq_args, &res.seq_res, 0);
- if (ret)
+ if (err)
goto out_free;

- /* Handle the case where the passed-in buffer is too short */
- if (res.acl_flags & NFS4_ACL_TRUNC) {
- /* Did the user only issue a request for the acl length? */
- if (buf == NULL)
- goto out_ok;
- ret = -ERANGE;
- goto out_free;
- }
- nfs4_write_cached_acl(inode, pages, res.acl_data_offset, res.acl_len);
- if (buf) {
- if (res.acl_len > buflen) {
- ret = -ERANGE;
- goto out_free;
- }
- _copy_from_pages(buf, pages, res.acl_data_offset, res.acl_len);
- }
-out_ok:
- ret = res.acl_len;
+ richacl_compute_max_masks(res.acl);
+ /* FIXME: Set inode->i_mode from res->mode? */
+ set_cached_richacl(inode, res.acl);
+ err = 0;
+
out_free:
- for (i = 0; i < npages; i++)
- if (pages[i])
- __free_page(pages[i]);
+ if (err) {
+ richacl_put(res.acl);
+ res.acl = ERR_PTR(err);
+ }
+ for (i = 0; i < ARRAY_SIZE(pages) && pages[i]; i++)
+ __free_page(pages[i]);
if (res.acl_scratch)
__free_page(res.acl_scratch);
- return ret;
+ return res.acl;
}

-static ssize_t nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t buflen)
+static struct richacl *nfs4_get_acl_uncached(struct inode *inode)
{
struct nfs4_exception exception = { };
- ssize_t ret;
+ struct richacl *acl;
do {
- ret = __nfs4_get_acl_uncached(inode, buf, buflen);
- trace_nfs4_get_acl(inode, ret);
- if (ret >= 0)
+ acl = __nfs4_get_acl_uncached(inode);
+ trace_nfs4_get_acl(inode, IS_ERR(acl) ? PTR_ERR(acl) : 0);
+ if (!IS_ERR(acl))
break;
- ret = nfs4_handle_exception(NFS_SERVER(inode), ret, &exception);
+ acl = ERR_PTR(nfs4_handle_exception(NFS_SERVER(inode),
+ PTR_ERR(acl), &exception));
} while (exception.retry);
- return ret;
+ return acl;
}

-static ssize_t nfs4_proc_get_acl(struct inode *inode, void *buf, size_t buflen)
+static struct richacl *nfs4_proc_get_acl(struct inode *inode)
{
struct nfs_server *server = NFS_SERVER(inode);
+ struct richacl *acl;
int ret;

- if (!nfs4_server_supports_acls(server))
- return -EOPNOTSUPP;
+ if (!(server->caps & (NFS_CAP_ALLOW_ACLS | NFS_CAP_DENY_ACLS)))
+ return ERR_PTR(-EOPNOTSUPP);
ret = nfs_revalidate_inode(server, inode);
if (ret < 0)
- return ret;
+ return ERR_PTR(ret);
if (NFS_I(inode)->cache_validity & NFS_INO_INVALID_ACL)
nfs_zap_acl_cache(inode);
- ret = nfs4_read_cached_acl(inode, buf, buflen);
- if (ret != -ENOENT)
- /* -ENOENT is returned if there is no ACL or if there is an ACL
- * but no cached acl data, just the acl length */
- return ret;
- return nfs4_get_acl_uncached(inode, buf, buflen);
+ acl = get_cached_richacl(inode);
+ if (acl != ACL_NOT_CACHED)
+ return acl;
+ return nfs4_get_acl_uncached(inode);
+}
+
+static int
+richacl_supported(struct nfs_server *server, struct richacl *acl)
+{
+ struct richace *ace;
+
+ if (!(server->caps & (NFS_CAP_ALLOW_ACLS | NFS_CAP_DENY_ACLS)))
+ return -EOPNOTSUPP;
+
+ richacl_for_each_entry(ace, acl) {
+ if (richace_is_allow(ace)) {
+ if (!(server->caps & NFS_CAP_ALLOW_ACLS))
+ return -EINVAL;
+ } else if (richace_is_deny(ace)) {
+ if (!(server->caps & NFS_CAP_DENY_ACLS))
+ return -EINVAL;
+ } else
+ return -EINVAL;
+ }
+ return 0;
}

-static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen)
+static int
+nfs4_encode_user(struct xdr_stream *xdr, const struct nfs_server *server,
+ kuid_t uid)
+{
+ char name[IDMAP_NAMESZ];
+ int len;
+ __be32 *p;
+
+ len = nfs_map_uid_to_name(server, uid, name, IDMAP_NAMESZ);
+ if (len < 0) {
+ dprintk("nfs: couldn't resolve uid %d to string\n",
+ from_kuid(&init_user_ns, uid));
+ return -ENOENT;
+ }
+ p = xdr_reserve_space(xdr, 4 + len);
+ if (!p)
+ return -EIO;
+ p = xdr_encode_opaque(p, name, len);
+ return 0;
+}
+
+static int
+nfs4_encode_group(struct xdr_stream *xdr, const struct nfs_server *server,
+ kgid_t gid)
+{
+ char name[IDMAP_NAMESZ];
+ int len;
+ __be32 *p;
+
+ len = nfs_map_gid_to_group(server, gid, name, IDMAP_NAMESZ);
+ if (len < 0) {
+ dprintk("nfs: couldn't resolve gid %d to string\n",
+ from_kgid(&init_user_ns, gid));
+ return -ENOENT;
+ }
+ p = xdr_reserve_space(xdr, 4 + len);
+ if (!p)
+ return -EIO;
+ p = xdr_encode_opaque(p, name, len);
+ return 0;
+}
+
+static unsigned int
+nfs4_ace_mask(int minorversion)
+{
+ return minorversion == 0 ? NFS40_ACE_MASK_ALL : NFS4_ACE_MASK_ALL;
+}
+
+static int
+nfs4_encode_ace_who(struct xdr_stream *xdr, const struct nfs_server *server,
+ struct richace *ace, struct richacl *acl)
+{
+ const char *who;
+ __be32 *p;
+
+ if (ace->e_flags & RICHACE_SPECIAL_WHO) {
+ unsigned int special_id = ace->e_id.special;
+ const char *who;
+ unsigned int len;
+
+ if (!nfs4acl_special_id_to_who(special_id, &who, &len)) {
+ WARN_ON_ONCE(1);
+ return -EIO;
+ }
+ p = xdr_reserve_space(xdr, 4 + len);
+ if (!p)
+ return -EIO;
+ xdr_encode_opaque(p, who, len);
+ return 0;
+ } else {
+ who = richace_unmapped_identifier(ace, acl);
+ if (who) {
+ unsigned int len = strlen(who);
+
+ p = xdr_reserve_space(xdr, 4 + len);
+ if (!p)
+ return -EIO;
+ xdr_encode_opaque(p, who, len);
+ return 0;
+ } else if (ace->e_flags & RICHACE_IDENTIFIER_GROUP)
+ return nfs4_encode_group(xdr, server, ace->e_id.gid);
+ else
+ return nfs4_encode_user(xdr, server, ace->e_id.uid);
+ }
+}
+
+static int
+nfs4_encode_acl(struct page **pages, unsigned int len, struct richacl *acl,
+ const struct nfs_server *server)
+{
+ int minorversion = server->nfs_client->cl_minorversion;
+ unsigned int ace_mask = nfs4_ace_mask(minorversion);
+ struct xdr_stream xdr;
+ struct xdr_buf buf;
+ __be32 *p;
+ struct richace *ace;
+
+ /* Reject acls not understood by the server */
+ if (server->attr_bitmask[1] & FATTR4_WORD1_DACL) {
+ BUILD_BUG_ON(NFS4_ACE_MASK_ALL != RICHACE_VALID_MASK);
+ } else {
+ if (acl->a_flags)
+ return -EINVAL;
+ richacl_for_each_entry(ace, acl) {
+ if (ace->e_flags & RICHACE_INHERITED_ACE)
+ return -EINVAL;
+ }
+ }
+ richacl_for_each_entry(ace, acl) {
+ if (ace->e_mask & ~ace_mask)
+ return -EINVAL;
+ }
+
+ xdr_init_encode_pages(&xdr, &buf, pages, len);
+
+ if (server->attr_bitmask[1] & FATTR4_WORD1_DACL) {
+ p = xdr_reserve_space(&xdr, 4);
+ if (!p)
+ goto fail;
+ *p = cpu_to_be32(acl ? acl->a_flags : 0);
+ }
+
+ p = xdr_reserve_space(&xdr, 4);
+ if (!p)
+ goto fail;
+ if (!acl) {
+ *p++ = cpu_to_be32(0);
+ return buf.len;
+ }
+ *p++ = cpu_to_be32(acl->a_count);
+
+ richacl_for_each_entry(ace, acl) {
+ p = xdr_reserve_space(&xdr, 4*3);
+ if (!p)
+ goto fail;
+ *p++ = cpu_to_be32(ace->e_type);
+ *p++ = cpu_to_be32(ace->e_flags &
+ ~(RICHACE_SPECIAL_WHO | RICHACE_UNMAPPED_WHO));
+ *p++ = cpu_to_be32(ace->e_mask & NFS4_ACE_MASK_ALL);
+ if (nfs4_encode_ace_who(&xdr, server, ace, acl) != 0)
+ goto fail;
+ }
+
+ return buf.len;
+
+fail:
+ return -ENOMEM;
+}
+
+static int __nfs4_proc_set_acl(struct inode *inode, struct richacl *acl)
{
struct nfs_server *server = NFS_SERVER(inode);
- struct page *pages[NFS4ACL_MAXPAGES];
+ struct page *pages[DIV_ROUND_UP(NFS4ACL_SIZE_MAX, PAGE_SIZE) + 1 /* scratch */] = {};
struct nfs_setaclargs arg = {
+ .server = server,
.fh = NFS_FH(inode),
.acl_pages = pages,
- .acl_len = buflen,
};
struct nfs_setaclres res;
struct rpc_message msg = {
@@ -4747,16 +4810,20 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
.rpc_argp = &arg,
.rpc_resp = &res,
};
- unsigned int npages = DIV_ROUND_UP(buflen, PAGE_SIZE);
int ret, i;

- if (!nfs4_server_supports_acls(server))
- return -EOPNOTSUPP;
- if (npages > ARRAY_SIZE(pages))
- return -ERANGE;
- i = buf_to_pages_noslab(buf, buflen, arg.acl_pages);
- if (i < 0)
- return i;
+ ret = richacl_supported(server, acl);
+ if (ret)
+ return ret;
+
+ ret = nfs4_encode_acl(pages, NFS4ACL_SIZE_MAX, acl, server);
+ if (ret < 0) {
+ for (i = 0; i < ARRAY_SIZE(pages) && pages[i]; i++)
+ put_page(pages[i]);
+ return ret;
+ }
+ arg.acl_len = ret;
+
nfs4_inode_return_delegation(inode);
ret = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1);

@@ -4764,8 +4831,8 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
* Free each page after tx, so the only ref left is
* held by the network stack
*/
- for (; i > 0; i--)
- put_page(pages[i-1]);
+ for (i = 0; i < ARRAY_SIZE(pages) && pages[i]; i++)
+ put_page(pages[i]);

/*
* Acl update can result in inode attribute update.
@@ -4779,12 +4846,12 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
return ret;
}

-static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen)
+static int nfs4_proc_set_acl(struct inode *inode, struct richacl *acl)
{
struct nfs4_exception exception = { };
int err;
do {
- err = __nfs4_proc_set_acl(inode, buf, buflen);
+ err = __nfs4_proc_set_acl(inode, acl);
trace_nfs4_set_acl(inode, err);
err = nfs4_handle_exception(NFS_SERVER(inode), err,
&exception);
@@ -6246,34 +6313,283 @@ nfs4_release_lockowner(struct nfs_server *server, struct nfs4_lock_state *lsp)
rpc_call_async(server->client, &msg, 0, &nfs4_release_lockowner_ops, data);
}

+static int nfs4_xattr_set_richacl(struct dentry *dentry, const char *key,
+ const void *buf, size_t buflen,
+ int flags, int handler_flags)
+{
+ struct inode *inode = d_inode(dentry);
+ struct richacl *acl;
+ int error;
+
+ if (strcmp(key, "") != 0)
+ return -EINVAL;
+
+ if (buf) {
+ acl = richacl_from_xattr(&init_user_ns, buf, buflen);
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+ error = richacl_apply_masks(&acl, inode->i_uid);
+ } else {
+ /*
+ * "Remove the acl"; only permissions granted by the mode
+ * remain. We are using the cached mode here which could be
+ * outdated; should we do a GETATTR first to narrow down the
+ * race window?
+ */
+ acl = richacl_from_mode(inode->i_mode);
+ error = 0;
+ }
+
+ if (!error)
+ error = nfs4_proc_set_acl(inode, acl);
+ richacl_put(acl);
+ return error;
+}
+
+static int nfs4_xattr_get_richacl(struct dentry *dentry, const char *key,
+ void *buf, size_t buflen, int handler_flags)
+{
+ struct inode *inode = d_inode(dentry);
+ struct richacl *acl;
+ int error;
+ mode_t mode = inode->i_mode & S_IFMT;
+
+ if (strcmp(key, "") != 0)
+ return -EINVAL;
+
+ acl = nfs4_proc_get_acl(inode);
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+ if (acl == NULL)
+ return -ENODATA;
+ error = -ENODATA;
+ if (richacl_equiv_mode(acl, &mode) == 0 &&
+ ((mode ^ inode->i_mode) & S_IRWXUGO) == 0)
+ goto out;
+ error = richacl_to_xattr(&init_user_ns, acl, buf, buflen);
+out:
+ richacl_put(acl);
+ return error;
+}
+
+static size_t nfs4_xattr_list_richacl(struct dentry *dentry, char *list,
+ size_t list_len, const char *name,
+ size_t name_len, int handler_flags)
+{
+ struct nfs_server *server = NFS_SERVER(d_inode(dentry));
+ size_t len = sizeof(XATTR_NAME_RICHACL);
+
+ if (!(server->caps & (NFS_CAP_ALLOW_ACLS | NFS_CAP_DENY_ACLS)))
+ return 0;
+
+ if (list && len <= list_len)
+ memcpy(list, XATTR_NAME_RICHACL, len);
+ return len;
+}
+
#define XATTR_NAME_NFSV4_ACL "system.nfs4_acl"

+static int richacl_to_nfs4_acl(struct nfs_server *server,
+ const struct richacl *acl,
+ void *buf, size_t buflen)
+{
+ const struct richace *ace;
+ __be32 *p = buf;
+ size_t size = 0;
+
+ size += sizeof(*p);
+ if (buflen >= size)
+ *p++ = cpu_to_be32(acl->a_count);
+
+ richacl_for_each_entry(ace, acl) {
+ char who_buf[IDMAP_NAMESZ];
+ const char *who = who_buf;
+ int who_len;
+
+ size += 3 * sizeof(*p);
+ if (buflen >= size) {
+ *p++ = cpu_to_be32(ace->e_type);
+ *p++ = cpu_to_be32(ace->e_flags &
+ ~(RICHACE_INHERITED_ACE |
+ RICHACE_UNMAPPED_WHO |
+ RICHACE_SPECIAL_WHO));
+ *p++ = cpu_to_be32(ace->e_mask);
+ }
+
+ if (richace_is_unix_user(ace)) {
+ who_len = nfs_map_uid_to_name(server, ace->e_id.uid,
+ who_buf, sizeof(who_buf));
+ if (who_len < 0)
+ return -EIO;
+ } else if (richace_is_unix_group(ace)) {
+ who_len = nfs_map_gid_to_group(server, ace->e_id.gid,
+ who_buf, sizeof(who_buf));
+ if (who_len < 0)
+ return -EIO;
+ } else if (ace->e_flags & RICHACE_SPECIAL_WHO) {
+ if (!nfs4acl_special_id_to_who(ace->e_id.special,
+ &who, &who_len))
+ return -EIO;
+ } else {
+ who = richace_unmapped_identifier(ace, acl);
+ if (who)
+ who_len = strlen(who);
+ else
+ return -EIO;
+ }
+
+ size += sizeof(*p) + ALIGN(who_len, sizeof(*p));
+ if (buflen >= size) {
+ unsigned int padding = -who_len & (sizeof(*p) - 1);
+
+ *p++ = cpu_to_be32(who_len);
+ memcpy(p, who, who_len);
+ memset((char *)p + who_len, 0, padding);
+ p += DIV_ROUND_UP(who_len, sizeof(*p));
+ }
+ }
+ if (buflen && buflen < size)
+ return -ERANGE;
+ return size;
+}
+
+static struct richacl *richacl_from_nfs4_acl(struct nfs_server *server,
+ const void *buf, size_t buflen)
+{
+ struct richacl *acl = NULL;
+ struct richace *ace;
+ const __be32 *p = buf;
+ int count, err;
+
+ if (buflen < sizeof(*p))
+ return ERR_PTR(-EINVAL);
+ count = be32_to_cpu(*p++);
+ if (count > RICHACL_XATTR_MAX_COUNT)
+ return ERR_PTR(-EINVAL);
+ buflen -= sizeof(*p);
+ acl = richacl_alloc(count, GFP_NOFS);
+ if (!acl)
+ return ERR_PTR(-ENOMEM);
+ richacl_for_each_entry(ace, acl) {
+ u32 who_len, size;
+ int special_id;
+ char *who;
+
+ err = -EINVAL;
+ if (buflen < 4 * sizeof(*p))
+ goto out;
+ ace->e_type = be32_to_cpu(*p++);
+ ace->e_flags = be32_to_cpu(*p++);
+ if (ace->e_flags & (RICHACE_SPECIAL_WHO | RICHACE_UNMAPPED_WHO))
+ goto out;
+ ace->e_mask = be32_to_cpu(*p++);
+ who_len = be32_to_cpu(*p++);
+ buflen -= 4 * sizeof(*p);
+ size = ALIGN(who_len, 4);
+ if (buflen < size || size == 0)
+ goto out;
+ who = (char *)p;
+ special_id = nfs4acl_who_to_special_id(who, who_len);
+ if (special_id >= 0) {
+ ace->e_flags |= RICHACE_SPECIAL_WHO;
+ ace->e_id.special = special_id;
+ } else {
+ bool unmappable;
+
+ if (ace->e_flags & RICHACE_IDENTIFIER_GROUP) {
+ err = nfs_map_group_to_gid(server, who, who_len,
+ &ace->e_id.gid);
+ if (err) {
+ dprintk("%s: nfs_map_group_to_gid "
+ "failed!\n", __func__);
+ goto out;
+ }
+ /* FIXME: nfsidmap doesn't distinguish between
+ group nobody and unmappable groups! */
+ unmappable = gid_eq(ace->e_id.gid,
+ make_kgid(&init_user_ns, 99));
+ } else {
+ err = nfs_map_name_to_uid(server, who, who_len,
+ &ace->e_id.uid);
+ if (err) {
+ dprintk("%s: nfs_map_name_to_gid "
+ "failed!\n", __func__);
+ goto out;
+ }
+ /* FIXME: nfsidmap doesn't distinguish between
+ user nobody and unmappable users! */
+ unmappable = uid_eq(ace->e_id.uid,
+ make_kuid(&init_user_ns, 99));
+ }
+ if (unmappable) {
+ err = -ENOMEM;
+ if (richacl_add_unmapped_identifier(&acl, &ace,
+ who, who_len, GFP_NOFS))
+ goto out;
+ }
+ }
+ p += size / sizeof(*p);
+ buflen -= size;
+ }
+ err = -EINVAL;
+ if (buflen != 0)
+ goto out;
+ err = 0;
+
+out:
+ if (err) {
+ richacl_put(acl);
+ acl = ERR_PTR(err);
+ }
+ return acl;
+}
+
static int nfs4_xattr_set_nfs4_acl(struct dentry *dentry, const char *key,
const void *buf, size_t buflen,
int flags, int type)
{
- if (strcmp(key, "") != 0)
+ struct inode *inode = d_inode(dentry);
+ struct richacl *acl;
+ int error;
+
+ if (!buf || strcmp(key, "") != 0)
return -EINVAL;

- return nfs4_proc_set_acl(d_inode(dentry), buf, buflen);
+ acl = richacl_from_nfs4_acl(NFS_SERVER(inode), (void *)buf, buflen);
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+ error = nfs4_proc_set_acl(inode, acl);
+ richacl_put(acl);
+ return error;
}

static int nfs4_xattr_get_nfs4_acl(struct dentry *dentry, const char *key,
void *buf, size_t buflen, int type)
{
+ struct inode *inode = d_inode(dentry);
+ struct richacl *acl;
+ int error;
+
if (strcmp(key, "") != 0)
return -EINVAL;
-
- return nfs4_proc_get_acl(d_inode(dentry), buf, buflen);
+ acl = nfs4_proc_get_acl(inode);
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+ if (acl == NULL)
+ return -ENODATA;
+ error = richacl_to_nfs4_acl(NFS_SERVER(inode), acl, buf, buflen);
+ richacl_put(acl);
+ return error;
}

static size_t nfs4_xattr_list_nfs4_acl(struct dentry *dentry, char *list,
size_t list_len, const char *name,
size_t name_len, int type)
{
+ struct nfs_server *server = NFS_SERVER(d_inode(dentry));
size_t len = sizeof(XATTR_NAME_NFSV4_ACL);

- if (!nfs4_server_supports_acls(NFS_SERVER(d_inode(dentry))))
+ if (!(server->caps & (NFS_CAP_ALLOW_ACLS | NFS_CAP_DENY_ACLS)))
return 0;

if (list && len <= list_len)
@@ -8826,6 +9142,13 @@ const struct nfs_rpc_ops nfs_v4_clientops = {
.clone_server = nfs_clone_server,
};

+static const struct xattr_handler nfs4_xattr_richacl_handler = {
+ .prefix = XATTR_NAME_RICHACL,
+ .list = nfs4_xattr_list_richacl,
+ .get = nfs4_xattr_get_richacl,
+ .set = nfs4_xattr_set_richacl,
+};
+
static const struct xattr_handler nfs4_xattr_nfs4_acl_handler = {
.prefix = XATTR_NAME_NFSV4_ACL,
.list = nfs4_xattr_list_nfs4_acl,
@@ -8834,6 +9157,7 @@ static const struct xattr_handler nfs4_xattr_nfs4_acl_handler = {
};

const struct xattr_handler *nfs4_xattr_handlers[] = {
+ &nfs4_xattr_richacl_handler,
&nfs4_xattr_nfs4_acl_handler,
#ifdef CONFIG_NFS_V4_SECURITY_LABEL
&nfs4_xattr_nfs4_label_handler,
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index eefed15..f2507d7 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -52,6 +52,10 @@
#include <linux/nfs.h>
#include <linux/nfs4.h>
#include <linux/nfs_fs.h>
+#include <linux/nfs_idmap.h>
+#include <linux/richacl.h>
+#include <linux/richacl_xattr.h> /* for RICHACL_XATTR_MAX_COUNT */
+#include <linux/nfs4acl.h>

#include "nfs4_fs.h"
#include "internal.h"
@@ -1650,16 +1654,24 @@ encode_restorefh(struct xdr_stream *xdr, struct compound_hdr *hdr)
static void
encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg, struct compound_hdr *hdr)
{
- __be32 *p;
+ int attrlen_offset;
+ __be32 attrlen, *p;

encode_op_hdr(xdr, OP_SETATTR, decode_setacl_maxsz, hdr);
encode_nfs4_stateid(xdr, &zero_stateid);
+
+ /* Encode attribute bitmap. */
p = reserve_space(xdr, 2*4);
*p++ = cpu_to_be32(1);
*p = cpu_to_be32(FATTR4_WORD0_ACL);
- p = reserve_space(xdr, 4);
- *p = cpu_to_be32(arg->acl_len);
+
+ attrlen_offset = xdr->buf->len;
+ xdr_reserve_space(xdr, 4); /* to be backfilled later */
+
xdr_write_pages(xdr, arg->acl_pages, 0, arg->acl_len);
+
+ attrlen = htonl(xdr->buf->len - attrlen_offset - 4);
+ write_bytes_to_xdr_buf(xdr->buf, attrlen_offset, &attrlen, 4);
}

static void
@@ -2488,7 +2500,7 @@ static void nfs4_xdr_enc_getacl(struct rpc_rqst *req, struct xdr_stream *xdr,
encode_sequence(xdr, &args->seq_args, &hdr);
encode_putfh(xdr, args->fh, &hdr);
replen = hdr.replen + op_decode_hdr_maxsz + 1;
- encode_getattr_two(xdr, FATTR4_WORD0_ACL, 0, &hdr);
+ encode_getattr_two(xdr, FATTR4_WORD0_ACL, FATTR4_WORD1_MODE, &hdr);

xdr_inline_pages(&req->rq_rcv_buf, replen << 2,
args->acl_pages, 0, args->acl_len);
@@ -5260,24 +5272,135 @@ decode_restorefh(struct xdr_stream *xdr)
return decode_op_hdr(xdr, OP_RESTOREFH);
}

+static int
+nfs4_decode_ace_who(struct richace *ace,
+ const char **unmapped, unsigned int *unmapped_len,
+ const struct nfs_server *server,
+ struct xdr_stream *xdr)
+{
+ char *who;
+ u32 len;
+ int special_id;
+ __be32 *p;
+ int error;
+
+ p = xdr_inline_decode(xdr, 4);
+ if (!p)
+ return -ENOMEM; /* acl truncated */
+ len = be32_to_cpup(p++);
+ if (len >= XDR_MAX_NETOBJ) {
+ dprintk("%s: name too long (%u)!\n",
+ __func__, len);
+ return -EIO;
+ }
+ who = (char *)xdr_inline_decode(xdr, len);
+ if (!who)
+ return -ENOMEM; /* acl truncated */
+
+ special_id = nfs4acl_who_to_special_id(who, len);
+ if (special_id >= 0) {
+ ace->e_flags |= RICHACE_SPECIAL_WHO;
+ ace->e_flags &= ~RICHACE_IDENTIFIER_GROUP;
+ ace->e_id.special = special_id;
+ return 0;
+ }
+ if (ace->e_flags & RICHACE_IDENTIFIER_GROUP) {
+ error = nfs_map_group_to_gid(server, who, len, &ace->e_id.gid);
+ if (error) {
+ dprintk("%s: nfs_map_group_to_gid failed!\n",
+ __func__);
+ return error;
+ }
+ /* FIXME: nfsidmap doesn't distinguish between group nobody and
+ unmappable groups! */
+ if (gid_eq(ace->e_id.gid, make_kgid(&init_user_ns, 99))) {
+ *unmapped = who;
+ *unmapped_len = len;
+ }
+ } else {
+ error = nfs_map_name_to_uid(server, who, len, &ace->e_id.uid);
+ if (error) {
+ dprintk("%s: nfs_map_name_to_uid failed!\n",
+ __func__);
+ return error;
+ }
+ /* FIXME: nfsidmap doesn't distinguish between user nobody and
+ unmappable users! */
+ if (uid_eq(ace->e_id.uid, make_kuid(&init_user_ns, 99))) {
+ *unmapped = who;
+ *unmapped_len = len;
+ }
+ }
+ return 0;
+}
+
+static struct richacl *
+decode_acl_entries(struct xdr_stream *xdr, const struct nfs_server *server)
+{
+ struct richacl *acl;
+ struct richace *ace;
+ uint32_t count;
+ __be32 *p;
+ int status;
+
+ p = xdr_inline_decode(xdr, 4);
+ if (unlikely(!p))
+ return ERR_PTR(-ENOMEM); /* acl truncated */
+ count = be32_to_cpup(p);
+ if (count > RICHACL_XATTR_MAX_COUNT)
+ return ERR_PTR(-EIO);
+ acl = richacl_alloc(count, GFP_NOFS);
+ if (!acl)
+ return ERR_PTR(-ENOMEM);
+ richacl_for_each_entry(ace, acl) {
+ const char *unmapped = NULL;
+ unsigned int unmapped_len;
+
+ p = xdr_inline_decode(xdr, 4*3);
+ status = -ENOMEM;
+ if (unlikely(!p))
+ goto out; /* acl truncated */
+ ace->e_type = be32_to_cpup(p++);
+ ace->e_flags = be32_to_cpup(p++);
+ status = -EIO;
+ if (ace->e_flags &
+ (RICHACE_SPECIAL_WHO | RICHACE_UNMAPPED_WHO))
+ goto out;
+ ace->e_mask = be32_to_cpup(p++);
+ status = nfs4_decode_ace_who(ace, &unmapped,
+ &unmapped_len, server,
+ xdr);
+ if (status)
+ goto out;
+ if (unmapped) {
+ status = -ENOMEM;
+ if (richacl_add_unmapped_identifier(&acl, &ace,
+ unmapped, unmapped_len,
+ GFP_NOFS))
+ goto out;
+ }
+ }
+ status = 0;
+
+out:
+ if (status) {
+ richacl_put(acl);
+ acl = ERR_PTR(status);
+ }
+ return acl;
+}
+
static int decode_getacl(struct xdr_stream *xdr, struct rpc_rqst *req,
struct nfs_getaclres *res)
{
unsigned int savep;
uint32_t attrlen,
bitmap[3] = {0};
+ struct richacl *acl = NULL;
int status;
- unsigned int pg_offset;

- res->acl_len = 0;
if ((status = decode_op_hdr(xdr, OP_GETATTR)) != 0)
goto out;
-
- xdr_enter_page(xdr, xdr->buf->page_len);
-
- /* Calculate the offset of the page data */
- pg_offset = xdr->buf->head[0].iov_len;
-
if ((status = decode_attr_bitmap(xdr, bitmap)) != 0)
goto out;
if ((status = decode_attr_length(xdr, &attrlen, &savep)) != 0)
@@ -5286,24 +5409,28 @@ static int decode_getacl(struct xdr_stream *xdr, struct rpc_rqst *req,
if (unlikely(bitmap[0] & (FATTR4_WORD0_ACL - 1U)))
return -EIO;
if (likely(bitmap[0] & FATTR4_WORD0_ACL)) {
-
- /* The bitmap (xdr len + bitmaps) and the attr xdr len words
- * are stored with the acl data to handle the problem of
- * variable length bitmaps.*/
- res->acl_data_offset = xdr_stream_pos(xdr) - pg_offset;
- res->acl_len = attrlen;
-
- /* Check for receive buffer overflow */
- if (res->acl_len > (xdr->nwords << 2) ||
- res->acl_len + res->acl_data_offset > xdr->buf->page_len) {
- res->acl_flags |= NFS4_ACL_TRUNC;
- dprintk("NFS: acl reply: attrlen %u > page_len %u\n",
- attrlen, xdr->nwords << 2);
- }
+ acl = decode_acl_entries(xdr, res->server);
+ status = PTR_ERR(acl);
+ if (IS_ERR(acl))
+ goto out;
+ bitmap[0] &= ~FATTR4_WORD0_ACL;
} else
status = -EOPNOTSUPP;

+ status = -EIO;
+ if (unlikely(bitmap[0]))
+ goto out;
+
+ status = decode_attr_mode(xdr, bitmap, &res->mode);
+ if (status < 0)
+ goto out;
+ status = 0;
+
out:
+ if (status == 0)
+ res->acl = acl;
+ else
+ richacl_put(acl);
return status;
}

diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 383a027..8ced33d 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -2319,7 +2319,7 @@ void nfs_fill_super(struct super_block *sb, struct nfs_mount_info *mount_info)
/* The VFS shouldn't apply the umask to mode bits. We will do
* so ourselves when necessary.
*/
- sb->s_flags |= MS_POSIXACL;
+ sb->s_flags |= MS_RICHACL;
sb->s_time_gran = 1;
}

@@ -2346,7 +2346,7 @@ void nfs_clone_super(struct super_block *sb, struct nfs_mount_info *mount_info)
/* The VFS shouldn't apply the umask to mode bits. We will do
* so ourselves when necessary.
*/
- sb->s_flags |= MS_POSIXACL;
+ sb->s_flags |= MS_RICHACL;
}

nfs_initialise_sb(sb);
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index c0e9614..b84e194 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -176,7 +176,6 @@ struct nfs_inode {
wait_queue_head_t waitqueue;

#if IS_ENABLED(CONFIG_NFS_V4)
- struct nfs4_cached_acl *nfs4_acl;
/* NFSv4 state */
struct list_head open_states;
struct nfs_delegation __rcu *delegation;
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 570a7df..6c41668 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -243,5 +243,7 @@ struct nfs_server {
#define NFS_CAP_ALLOCATE (1U << 20)
#define NFS_CAP_DEALLOCATE (1U << 21)
#define NFS_CAP_LAYOUTSTATS (1U << 22)
+#define NFS_CAP_ALLOW_ACLS (1U << 23)
+#define NFS_CAP_DENY_ACLS (1U << 24)

#endif
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 090ade4..337c341 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -683,9 +683,10 @@ struct nfs_setattrargs {

struct nfs_setaclargs {
struct nfs4_sequence_args seq_args;
+ const struct nfs_server * server;
struct nfs_fh * fh;
- size_t acl_len;
struct page ** acl_pages;
+ size_t acl_len;
};

struct nfs_setaclres {
@@ -703,9 +704,9 @@ struct nfs_getaclargs {
#define NFS4_ACL_TRUNC 0x0001 /* ACL was truncated */
struct nfs_getaclres {
struct nfs4_sequence_res seq_res;
- size_t acl_len;
- size_t acl_data_offset;
- int acl_flags;
+ const struct nfs_server * server;
+ struct richacl * acl;
+ umode_t mode;
struct page * acl_scratch;
};

--
2.4.3


2015-09-27 22:13:23

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 38/41] nfs: Remove unused xdr page offsets in getacl/setacl arguments

The arguments passed around for getacl and setacl xdr encoding, struct
nfs_setaclargs and struct nfs_getaclargs, both contain an array of
pages, an offset into the first page, and the length of the page data.
The offset is unused as it is always zero; remove it.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/nfs/nfs4proc.c | 5 ++---
fs/nfs/nfs4xdr.c | 4 ++--
include/linux/nfs_xdr.h | 2 --
3 files changed, 4 insertions(+), 7 deletions(-)

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index f93b9cd..b3a6558 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -4519,7 +4519,7 @@ static inline int nfs4_server_supports_acls(struct nfs_server *server)
#define NFS4ACL_MAXPAGES DIV_ROUND_UP(XATTR_SIZE_MAX, PAGE_SIZE)

static int buf_to_pages_noslab(const void *buf, size_t buflen,
- struct page **pages, unsigned int *pgbase)
+ struct page **pages)
{
struct page *newpage, **spages;
int rc = 0;
@@ -4663,7 +4663,6 @@ static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t bu
goto out_free;

args.acl_len = npages * PAGE_SIZE;
- args.acl_pgbase = 0;

dprintk("%s buf %p buflen %zu npages %d args.acl_len %zu\n",
__func__, buf, buflen, npages, args.acl_len);
@@ -4755,7 +4754,7 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
return -EOPNOTSUPP;
if (npages > ARRAY_SIZE(pages))
return -ERANGE;
- i = buf_to_pages_noslab(buf, buflen, arg.acl_pages, &arg.acl_pgbase);
+ i = buf_to_pages_noslab(buf, buflen, arg.acl_pages);
if (i < 0)
return i;
nfs4_inode_return_delegation(inode);
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 6f6d921..eefed15 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -1659,7 +1659,7 @@ encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg, struct compoun
*p = cpu_to_be32(FATTR4_WORD0_ACL);
p = reserve_space(xdr, 4);
*p = cpu_to_be32(arg->acl_len);
- xdr_write_pages(xdr, arg->acl_pages, arg->acl_pgbase, arg->acl_len);
+ xdr_write_pages(xdr, arg->acl_pages, 0, arg->acl_len);
}

static void
@@ -2491,7 +2491,7 @@ static void nfs4_xdr_enc_getacl(struct rpc_rqst *req, struct xdr_stream *xdr,
encode_getattr_two(xdr, FATTR4_WORD0_ACL, 0, &hdr);

xdr_inline_pages(&req->rq_rcv_buf, replen << 2,
- args->acl_pages, args->acl_pgbase, args->acl_len);
+ args->acl_pages, 0, args->acl_len);

encode_nops(&hdr);
}
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 52faf7e..090ade4 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -685,7 +685,6 @@ struct nfs_setaclargs {
struct nfs4_sequence_args seq_args;
struct nfs_fh * fh;
size_t acl_len;
- unsigned int acl_pgbase;
struct page ** acl_pages;
};

@@ -697,7 +696,6 @@ struct nfs_getaclargs {
struct nfs4_sequence_args seq_args;
struct nfs_fh * fh;
size_t acl_len;
- unsigned int acl_pgbase;
struct page ** acl_pages;
};

--
2.4.3


2015-09-27 22:13:18

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 37/41] nfs: Fix GETATTR bitmap verification

When decoding GETATTR replies, the client checks the attribute bitmap
for which attributes the server has sent. It misses bits at the word
boundaries, though; fix that.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/nfs/nfs4xdr.c | 23 +++++++++++++++++++++++
1 file changed, 23 insertions(+)

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index 788adf3..6f6d921 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -4375,6 +4375,11 @@ static int decode_statfs(struct xdr_stream *xdr, struct nfs_fsstat *fsstat)
goto xdr_error;
if ((status = decode_attr_files_total(xdr, bitmap, &fsstat->tfiles)) != 0)
goto xdr_error;
+
+ status = -EIO;
+ if (unlikely(bitmap[0]))
+ goto xdr_error;
+
if ((status = decode_attr_space_avail(xdr, bitmap, &fsstat->abytes)) != 0)
goto xdr_error;
if ((status = decode_attr_space_free(xdr, bitmap, &fsstat->fbytes)) != 0)
@@ -4574,6 +4579,10 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap,
goto xdr_error;
fattr->valid |= status;

+ status = -EIO;
+ if (unlikely(bitmap[0]))
+ goto xdr_error;
+
status = decode_attr_mode(xdr, bitmap, &fmode);
if (status < 0)
goto xdr_error;
@@ -4627,6 +4636,10 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap,
goto xdr_error;
fattr->valid |= status;

+ status = -EIO;
+ if (unlikely(bitmap[1]))
+ goto xdr_error;
+
status = decode_attr_mdsthreshold(xdr, bitmap, fattr->mdsthreshold);
if (status < 0)
goto xdr_error;
@@ -4789,12 +4802,22 @@ static int decode_fsinfo(struct xdr_stream *xdr, struct nfs_fsinfo *fsinfo)
if ((status = decode_attr_maxwrite(xdr, bitmap, &fsinfo->wtmax)) != 0)
goto xdr_error;
fsinfo->wtpref = fsinfo->wtmax;
+
+ status = -EIO;
+ if (unlikely(bitmap[0]))
+ goto xdr_error;
+
status = decode_attr_time_delta(xdr, bitmap, &fsinfo->time_delta);
if (status != 0)
goto xdr_error;
status = decode_attr_pnfstype(xdr, bitmap, &fsinfo->layouttype);
if (status != 0)
goto xdr_error;
+
+ status = -EIO;
+ if (unlikely(bitmap[1]))
+ goto xdr_error;
+
status = decode_attr_layout_blksize(xdr, bitmap, &fsinfo->blksize);
if (status)
goto xdr_error;
--
2.4.3


2015-09-27 22:13:12

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 36/41] sunrpc: Add xdr_init_encode_pages

Initialize xdr_stream and xdr_buf from a pages array, for encoding into
the pages.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
include/linux/sunrpc/xdr.h | 2 ++
net/sunrpc/xdr.c | 25 +++++++++++++++++++++++++
2 files changed, 27 insertions(+)

diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h
index 70c6b92..2c99cff 100644
--- a/include/linux/sunrpc/xdr.h
+++ b/include/linux/sunrpc/xdr.h
@@ -214,6 +214,8 @@ typedef void (*kxdreproc_t)(void *rqstp, struct xdr_stream *xdr, void *obj);
typedef int (*kxdrdproc_t)(void *rqstp, struct xdr_stream *xdr, void *obj);

extern void xdr_init_encode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p);
+extern void xdr_init_encode_pages(struct xdr_stream *xdr, struct xdr_buf *buf,
+ struct page **pages, unsigned int len);
extern __be32 *xdr_reserve_space(struct xdr_stream *xdr, size_t nbytes);
extern void xdr_commit_encode(struct xdr_stream *xdr);
extern void xdr_truncate_encode(struct xdr_stream *xdr, size_t len);
diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c
index 63c1c36..f97b96b 100644
--- a/net/sunrpc/xdr.c
+++ b/net/sunrpc/xdr.c
@@ -483,6 +483,31 @@ void xdr_init_encode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p)
EXPORT_SYMBOL_GPL(xdr_init_encode);

/**
+ * xdr_init_encode_pages - Initialize struct xdr_stream for encoding into pages
+ * @xdr: pointer to xdr_stream struct
+ * @buf: pointer to XDR buffer in which to encode data
+ * @pages: pages array in which to encode
+ * @len: maximum length of @buf
+ *
+ * Initialize @xdr and @buf for encoding into the @pages array. If a
+ * page in @pages is NULL, it will be allocated on demand.
+ */
+void xdr_init_encode_pages(struct xdr_stream *xdr, struct xdr_buf *buf,
+ struct page **pages, unsigned int len)
+{
+ memset(buf, 0, sizeof(*buf));
+ buf->pages = pages;
+ buf->page_len = len;
+ buf->buflen = len;
+
+ memset(xdr, 0, sizeof(*xdr));
+ xdr->buf = buf;
+ xdr->iov = buf->head;
+ xdr->page_ptr = pages - 1;
+}
+EXPORT_SYMBOL_GPL(xdr_init_encode_pages);
+
+/**
* xdr_commit_encode - Ensure all data is written to buffer
* @xdr: pointer to xdr_stream
*
--
2.4.3


2015-09-27 22:13:06

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 35/41] sunrpc: Allow to demand-allocate pages to encode into

When encoding large, variable-length objects such as acls into xdr_bufs,
it is easier to allocate buffer pages on demand rather than precomputing
the required buffer size.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
net/sunrpc/xdr.c | 9 +++++++++
1 file changed, 9 insertions(+)

diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c
index 4439ac4..63c1c36 100644
--- a/net/sunrpc/xdr.c
+++ b/net/sunrpc/xdr.c
@@ -537,6 +537,15 @@ static __be32 *xdr_get_next_encode_buffer(struct xdr_stream *xdr,
*/
xdr->scratch.iov_base = xdr->p;
xdr->scratch.iov_len = frag1bytes;
+
+ if (!*xdr->page_ptr) {
+ struct page *page = alloc_page(GFP_NOFS);
+
+ if (!page)
+ return NULL;
+ *xdr->page_ptr = page;
+ }
+
p = page_address(*xdr->page_ptr);
/*
* Note this is where the next encode will start after we've
--
2.4.3


2015-09-27 22:13:00

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 34/41] ext4: Don't allow unmapped identifiers in richacls

Don't allow acls which contain unmapped identifiers: they are meaningful
for remote file systems only.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/ext4/richacl.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/fs/ext4/richacl.c b/fs/ext4/richacl.c
index 6758def..ca6e0cb 100644
--- a/fs/ext4/richacl.c
+++ b/fs/ext4/richacl.c
@@ -68,8 +68,13 @@ ext4_set_richacl(handle_t *handle, struct inode *inode, struct richacl *acl)
int retval;

if (acl) {
- mode_t mode = inode->i_mode;
+ mode_t mode;

+ /* Don't allow acls with unmapped identifiers. */
+ if (richacl_has_unmapped_identifiers(acl))
+ return -EINVAL;
+
+ mode = inode->i_mode;
if (richacl_equiv_mode(acl, &mode) == 0) {
inode->i_mode = mode;
ext4_mark_inode_dirty(handle, inode);
--
2.4.3


2015-09-27 22:12:55

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 33/41] richacl: Add support for unmapped identifiers

Some remote file systems like nfs may return user or group identifiers
that cannot be mapped to local uids / gids. Allow to represent such
unmapped identifiers in richacls. (We still cannot represent unmapped
owners and owning groups, however.)

In the in-memory representation, the richacl is followed by a list of
NUL-terminated strings, with no padding. Entries with an unmapped
identifier have the RICHACE_UNMAPPED_WHO flag set, and ace->e_id.offs
specifies the offset into this list. Multiple entries can refer to the
same offset.

The xattr representation is similar, but ace->e_id is ignored, and the
list of unmapped identifier strings contains a string for each acl entry
whose RICHACE_UNMAPPED_WHO flag is set.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/richacl_base.c | 139 ++++++++++++++++++++++++++++++++++++++++++++----
fs/richacl_compat.c | 18 +++----
fs/richacl_inode.c | 4 +-
fs/richacl_xattr.c | 69 ++++++++++++++++++++----
include/linux/richacl.h | 33 ++++++++++--
5 files changed, 227 insertions(+), 36 deletions(-)

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index e0357ae..b1b3f5d 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -23,22 +23,25 @@
MODULE_LICENSE("GPL");

/**
- * richacl_alloc - allocate a richacl
+ * __richacl_alloc - allocate a richacl
* @count: number of entries
+ * @unmapped_size: size to reserve for unmapped identifiers
*/
struct richacl *
-richacl_alloc(int count, gfp_t gfp)
+__richacl_alloc(unsigned int count, size_t unmapped_size, gfp_t gfp)
{
- size_t size = sizeof(struct richacl) + count * sizeof(struct richace);
+ size_t size = sizeof(struct richacl) + count * sizeof(struct richace) +
+ unmapped_size;
struct richacl *acl = kzalloc(size, gfp);

if (acl) {
atomic_set(&acl->a_base.ba_refcount, 1);
acl->a_count = count;
+ acl->a_unmapped_size = unmapped_size;
}
return acl;
}
-EXPORT_SYMBOL_GPL(richacl_alloc);
+EXPORT_SYMBOL_GPL(__richacl_alloc);

/**
* richacl_clone - create a copy of a richacl
@@ -47,7 +50,8 @@ struct richacl *
richacl_clone(const struct richacl *acl, gfp_t gfp)
{
int count = acl->a_count;
- size_t size = sizeof(struct richacl) + count * sizeof(struct richace);
+ size_t size = sizeof(struct richacl) + count * sizeof(struct richace) +
+ acl->a_unmapped_size;
struct richacl *dup = kmalloc(size, gfp);

if (dup) {
@@ -59,6 +63,9 @@ richacl_clone(const struct richacl *acl, gfp_t gfp)

/**
* richace_copy - copy an acl entry
+ *
+ * If @from has an unmapped who value (from->e_flags & RICHACE_UNMAPPED_WHO),
+ * it can only be copied within the same acl!
*/
void
richace_copy(struct richace *to, const struct richace *from)
@@ -66,6 +73,82 @@ richace_copy(struct richace *to, const struct richace *from)
memcpy(to, from, sizeof(struct richace));
}

+/**
+ * richacl_add_unmapped_identifier
+ * @pacl: Pointer to an acl
+ * @pace: acl entry within @acl
+ * @who: unmapped identifier
+ * @len: length of @who
+ * @gfp: memory allocation flags
+ *
+ * Add an unmapped identifier to an acl, possibly reallocating the acl.
+ */
+int richacl_add_unmapped_identifier(struct richacl **pacl,
+ struct richace **pace,
+ const char *who,
+ unsigned int len, gfp_t gfp)
+{
+ struct richacl *acl = *pacl;
+ size_t size = sizeof(struct richacl) +
+ acl->a_count * sizeof(struct richace) +
+ acl->a_unmapped_size + len + 1;
+ unsigned int index = *pace - acl->a_entries;
+
+ acl = krealloc(*pacl, size, gfp);
+ if (acl) {
+ char *unmapped = (char *)(acl->a_entries + acl->a_count);
+ struct richace *ace = acl->a_entries + index;
+
+ ace->e_flags |= RICHACE_UNMAPPED_WHO;
+ ace->e_flags &= ~RICHACE_SPECIAL_WHO;
+ ace->e_id.offs = acl->a_unmapped_size;
+ memcpy(unmapped + ace->e_id.offs, who, len);
+ unmapped[ace->e_id.offs + len] = 0;
+ acl->a_unmapped_size += len + 1;
+ *pace = ace;
+ *pacl = acl;
+ return 0;
+ }
+ return -1;
+}
+EXPORT_SYMBOL_GPL(richacl_add_unmapped_identifier);
+
+/**
+ * richace_unmapped_identifier - get unmapped identifier
+ * @acl: acl containing @ace
+ * @ace: acl entry
+ *
+ * Get the unmapped identifier of @ace as a NUL-terminated string, or NULL if
+ * @ace doesn't have an unmapped identifier.
+ */
+const char *richace_unmapped_identifier(const struct richace *ace,
+ const struct richacl *acl)
+{
+ const char *unmapped = (char *)(acl->a_entries + acl->a_count);
+
+ if (!(ace->e_flags & RICHACE_UNMAPPED_WHO))
+ return NULL;
+ return unmapped + ace->e_id.offs;
+}
+EXPORT_SYMBOL(richace_unmapped_identifier);
+
+/**
+ * richacl_has_unmapped_identifiers
+ *
+ * Check if an acl has unmapped identifiers.
+ */
+bool richacl_has_unmapped_identifiers(struct richacl *acl)
+{
+ struct richace *ace;
+
+ richacl_for_each_entry(ace, acl) {
+ if (ace->e_flags & RICHACE_UNMAPPED_WHO)
+ return true;
+ }
+ return false;
+}
+EXPORT_SYMBOL_GPL(richacl_has_unmapped_identifiers);
+
/*
* richacl_mask_to_mode - compute the file permission bits from mask
* @mask: %RICHACE_* permission mask
@@ -213,7 +296,7 @@ static unsigned int richacl_allowed_to_who(struct richacl *acl,
richacl_for_each_entry_reverse(ace, acl) {
if (richace_is_inherit_only(ace))
continue;
- if (richace_is_same_identifier(ace, who) ||
+ if (richace_is_same_identifier(acl, ace, who) ||
richace_is_everyone(ace)) {
if (richace_is_allow(ace))
allowed |= ace->e_mask;
@@ -502,45 +585,72 @@ richacl_inherit(const struct richacl *dir_acl, int isdir)
const struct richace *dir_ace;
struct richacl *acl = NULL;
struct richace *ace;
- int count = 0;
+ unsigned int count = 0, unmapped_size = 0, offset = 0;
+ const char *dir_unmapped;
+ char *unmapped;

if (isdir) {
richacl_for_each_entry(dir_ace, dir_acl) {
if (!richace_is_inheritable(dir_ace))
continue;
+
count++;
+ dir_unmapped =
+ richace_unmapped_identifier(dir_ace, dir_acl);
+ if (dir_unmapped)
+ unmapped_size += strlen(dir_unmapped) + 1;
}
if (!count)
return NULL;
- acl = richacl_alloc(count, GFP_KERNEL);
+ acl = __richacl_alloc(count, unmapped_size, GFP_KERNEL);
if (!acl)
return ERR_PTR(-ENOMEM);
ace = acl->a_entries;
+ unmapped = (char *)(acl->a_entries + acl->a_count);
richacl_for_each_entry(dir_ace, dir_acl) {
if (!richace_is_inheritable(dir_ace))
continue;
+
richace_copy(ace, dir_ace);
if (dir_ace->e_flags & RICHACE_NO_PROPAGATE_INHERIT_ACE)
ace->e_flags &= ~RICHACE_INHERITANCE_FLAGS;
else if (!(dir_ace->e_flags & RICHACE_DIRECTORY_INHERIT_ACE))
ace->e_flags |= RICHACE_INHERIT_ONLY_ACE;
+
+ dir_unmapped =
+ richace_unmapped_identifier(dir_ace, dir_acl);
+ if (dir_unmapped) {
+ size_t sz = strlen(dir_unmapped) + 1;
+
+ ace->e_id.offs = offset;
+ memcpy(unmapped, dir_unmapped, sz);
+ unmapped += sz;
+ offset += sz;
+ }
ace++;
}
} else {
richacl_for_each_entry(dir_ace, dir_acl) {
if (!(dir_ace->e_flags & RICHACE_FILE_INHERIT_ACE))
continue;
+
count++;
+ dir_unmapped =
+ richace_unmapped_identifier(dir_ace, dir_acl);
+ if (dir_unmapped)
+ unmapped_size += strlen(dir_unmapped) + 1;
}
if (!count)
return NULL;
- acl = richacl_alloc(count, GFP_KERNEL);
+ acl = __richacl_alloc(count, unmapped_size, GFP_KERNEL);
if (!acl)
return ERR_PTR(-ENOMEM);
ace = acl->a_entries;
+ unmapped = (char *)(acl->a_entries + acl->a_count);
richacl_for_each_entry(dir_ace, dir_acl) {
if (!(dir_ace->e_flags & RICHACE_FILE_INHERIT_ACE))
continue;
+
richace_copy(ace, dir_ace);
ace->e_flags &= ~RICHACE_INHERITANCE_FLAGS;
/*
@@ -548,6 +658,17 @@ richacl_inherit(const struct richacl *dir_acl, int isdir)
* non-directories, so clear it.
*/
ace->e_mask &= ~RICHACE_DELETE_CHILD;
+
+ dir_unmapped =
+ richace_unmapped_identifier(dir_ace, dir_acl);
+ if (dir_unmapped) {
+ size_t sz = strlen(dir_unmapped) + 1;
+
+ ace->e_id.offs = offset;
+ memcpy(unmapped, dir_unmapped, sz);
+ unmapped += sz;
+ offset += sz;
+ }
ace++;
}
}
diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index df807c0..39805bc 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -71,11 +71,13 @@ richacl_insert_entry(struct richacl_alloc *alloc, struct richace **ace)
{
struct richacl *acl = alloc->acl;
unsigned int index = *ace - acl->a_entries;
- size_t tail_size = (acl->a_count - index) * sizeof(struct richace);
+ size_t tail_size = (acl->a_count - index) * sizeof(struct richace) +
+ acl->a_unmapped_size;

if (alloc->count == acl->a_count) {
size_t new_size = sizeof(struct richacl) +
- (acl->a_count + 1) * sizeof(struct richace);
+ (acl->a_count + 1) * sizeof(struct richace) +
+ acl->a_unmapped_size;

acl = krealloc(acl, new_size, GFP_KERNEL);
if (!acl)
@@ -103,10 +105,6 @@ struct richace *richacl_append_entry(struct richacl_alloc *alloc)
struct richacl *acl = alloc->acl;
struct richace *ace = acl->a_entries + acl->a_count;

- if (alloc->count > alloc->acl->a_count) {
- acl->a_count++;
- return ace;
- }
return richacl_insert_entry(alloc, &ace) ? NULL : ace;
}
EXPORT_SYMBOL_GPL(richacl_append_entry);
@@ -261,12 +259,12 @@ __richacl_propagate_everyone(struct richacl_alloc *alloc, struct richace *who,
if (richace_is_inherit_only(ace))
continue;
if (richace_is_allow(ace)) {
- if (richace_is_same_identifier(ace, who)) {
+ if (richace_is_same_identifier(acl, ace, who)) {
allow &= ~ace->e_mask;
allow_last = ace;
}
} else if (richace_is_deny(ace)) {
- if (richace_is_same_identifier(ace, who))
+ if (richace_is_same_identifier(acl, ace, who))
allow &= ~ace->e_mask;
else if (allow & ace->e_mask)
allow_last = NULL;
@@ -613,7 +611,7 @@ __richacl_isolate_who(struct richacl_alloc *alloc, struct richace *who,
richacl_for_each_entry(ace, acl) {
if (richace_is_inherit_only(ace))
continue;
- if (richace_is_same_identifier(ace, who))
+ if (richace_is_same_identifier(acl, ace, who))
deny &= ~ace->e_mask;
}
if (!deny)
@@ -629,7 +627,7 @@ __richacl_isolate_who(struct richacl_alloc *alloc, struct richace *who,
if (richace_is_inherit_only(ace))
continue;
if (richace_is_deny(ace)) {
- if (richace_is_same_identifier(ace, who))
+ if (richace_is_same_identifier(acl, ace, who))
return richace_change_mask(alloc, &ace,
ace->e_mask | deny);
} else if (richace_is_allow(ace) &&
diff --git a/fs/richacl_inode.c b/fs/richacl_inode.c
index d451e83..5d35346 100644
--- a/fs/richacl_inode.c
+++ b/fs/richacl_inode.c
@@ -160,8 +160,10 @@ richacl_permission(struct inode *inode, const struct richacl *acl,
} else if (richace_is_unix_group(ace)) {
if (!in_group_p(ace->e_id.gid))
continue;
- } else
+ } else if (richace_is_everyone(ace))
goto entry_matches_everyone;
+ else
+ continue;

/*
* Apply the group file mask to entries other than owner@ and
diff --git a/fs/richacl_xattr.c b/fs/richacl_xattr.c
index cd9979d..e9622f6 100644
--- a/fs/richacl_xattr.c
+++ b/fs/richacl_xattr.c
@@ -33,7 +33,8 @@ richacl_from_xattr(struct user_namespace *user_ns,
const struct richace_xattr *xattr_ace = (void *)(xattr_acl + 1);
struct richacl *acl;
struct richace *ace;
- int count;
+ unsigned int count, offset;
+ char *unmapped;

if (size < sizeof(*xattr_acl) ||
xattr_acl->a_version != RICHACL_XATTR_VERSION ||
@@ -43,10 +44,11 @@ richacl_from_xattr(struct user_namespace *user_ns,
count = le16_to_cpu(xattr_acl->a_count);
if (count > RICHACL_XATTR_MAX_COUNT)
return ERR_PTR(-EINVAL);
- if (size != count * sizeof(*xattr_ace))
+ if (size < count * sizeof(*xattr_ace))
return ERR_PTR(-EINVAL);
+ size -= count * sizeof(*xattr_ace);

- acl = richacl_alloc(count, GFP_NOFS);
+ acl = __richacl_alloc(count, size, GFP_NOFS);
if (!acl)
return ERR_PTR(-ENOMEM);

@@ -61,6 +63,16 @@ richacl_from_xattr(struct user_namespace *user_ns,
if (acl->a_other_mask & ~RICHACE_VALID_MASK)
goto fail_einval;

+ unmapped = (char *)(acl->a_entries + count);
+ if (size) {
+ char *xattr_unmapped = (char *)(xattr_ace + count);
+
+ if (xattr_unmapped[size - 1] != 0)
+ goto fail_einval;
+ memcpy(unmapped, xattr_unmapped, size);
+ }
+ offset = 0;
+
richacl_for_each_entry(ace, acl) {
ace->e_type = le16_to_cpu(xattr_ace->e_type);
ace->e_flags = le16_to_cpu(xattr_ace->e_flags);
@@ -72,6 +84,15 @@ richacl_from_xattr(struct user_namespace *user_ns,
ace->e_id.special = le32_to_cpu(xattr_ace->e_id);
if (ace->e_id.special > RICHACE_EVERYONE_SPECIAL_ID)
goto fail_einval;
+ } else if (ace->e_flags & RICHACE_UNMAPPED_WHO) {
+ size_t sz;
+
+ if (offset == size)
+ goto fail_einval;
+ ace->e_id.offs = offset;
+ sz = strlen(unmapped) + 1;
+ unmapped += sz;
+ offset += sz;
} else if (ace->e_flags & RICHACE_IDENTIFIER_GROUP) {
u32 id = le32_to_cpu(xattr_ace->e_id);

@@ -88,10 +109,12 @@ richacl_from_xattr(struct user_namespace *user_ns,
if (ace->e_type > RICHACE_ACCESS_DENIED_ACE_TYPE ||
(ace->e_mask & ~RICHACE_VALID_MASK))
goto fail_einval;
-
xattr_ace++;
}

+ if (offset != size)
+ goto fail_einval;
+
return acl;

fail_einval:
@@ -107,8 +130,15 @@ size_t
richacl_xattr_size(const struct richacl *acl)
{
size_t size = sizeof(struct richacl_xattr);
+ const struct richace *ace;

size += sizeof(struct richace_xattr) * acl->a_count;
+ richacl_for_each_entry(ace, acl) {
+ const char *unmapped = richace_unmapped_identifier(ace, acl);
+
+ if (unmapped)
+ size += strlen(unmapped) + 1;
+ }
return size;
}
EXPORT_SYMBOL_GPL(richacl_xattr_size);
@@ -127,6 +157,7 @@ richacl_to_xattr(struct user_namespace *user_ns,
struct richace_xattr *xattr_ace;
const struct richace *ace;
size_t real_size;
+ char *xattr_unmapped;

real_size = richacl_xattr_size(acl);
if (!buffer)
@@ -143,18 +174,33 @@ richacl_to_xattr(struct user_namespace *user_ns,
xattr_acl->a_other_mask = cpu_to_le32(acl->a_other_mask);

xattr_ace = (void *)(xattr_acl + 1);
+ xattr_unmapped = (char *)(xattr_ace + acl->a_count);
richacl_for_each_entry(ace, acl) {
+ const char *who;
+
xattr_ace->e_type = cpu_to_le16(ace->e_type);
xattr_ace->e_flags = cpu_to_le16(ace->e_flags);
xattr_ace->e_mask = cpu_to_le32(ace->e_mask);
if (ace->e_flags & RICHACE_SPECIAL_WHO)
xattr_ace->e_id = cpu_to_le32(ace->e_id.special);
- else if (ace->e_flags & RICHACE_IDENTIFIER_GROUP)
- xattr_ace->e_id =
- cpu_to_le32(from_kgid(user_ns, ace->e_id.gid));
- else
- xattr_ace->e_id =
- cpu_to_le32(from_kuid(user_ns, ace->e_id.uid));
+ else {
+ who = richace_unmapped_identifier(ace, acl);
+ if (who) {
+ size_t sz = strlen(who) + 1;
+
+ xattr_ace->e_id = 0;
+ memcpy(xattr_unmapped, who, sz);
+ xattr_unmapped += sz;
+ } else if (ace->e_flags & RICHACE_IDENTIFIER_GROUP) {
+ u32 id = from_kgid(user_ns, ace->e_id.gid);
+
+ xattr_ace->e_id = cpu_to_le32(id);
+ } else {
+ u32 id = from_kuid(user_ns, ace->e_id.uid);
+
+ xattr_ace->e_id = cpu_to_le32(id);
+ }
+ }
xattr_ace++;
}
return real_size;
@@ -184,7 +230,8 @@ static void richacl_fix_xattr_userns(
return;
count = size / sizeof(*xattr_ace);
for (; count; count--, xattr_ace++) {
- if (xattr_ace->e_flags & cpu_to_le16(RICHACE_SPECIAL_WHO))
+ if (xattr_ace->e_flags & cpu_to_le16(RICHACE_SPECIAL_WHO |
+ RICHACE_UNMAPPED_WHO))
continue;
if (xattr_ace->e_flags &
cpu_to_le16(RICHACE_IDENTIFIER_GROUP)) {
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index f48c04e..1822666 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -29,6 +29,7 @@ struct richace {
kuid_t uid;
kgid_t gid;
unsigned int special;
+ unsigned short offs; /* unmapped offset */
} e_id;
};

@@ -39,6 +40,7 @@ struct richacl {
unsigned int a_other_mask;
unsigned short a_count;
unsigned short a_flags;
+ unsigned short a_unmapped_size;
struct richace a_entries[0];
};

@@ -77,6 +79,7 @@ struct richacl {
#define RICHACE_INHERIT_ONLY_ACE 0x0008
#define RICHACE_IDENTIFIER_GROUP 0x0040
#define RICHACE_INHERITED_ACE 0x0080
+#define RICHACE_UNMAPPED_WHO 0x2000
#define RICHACE_SPECIAL_WHO 0x4000

#define RICHACE_VALID_FLAGS ( \
@@ -86,6 +89,7 @@ struct richacl {
RICHACE_INHERIT_ONLY_ACE | \
RICHACE_IDENTIFIER_GROUP | \
RICHACE_INHERITED_ACE | \
+ RICHACE_UNMAPPED_WHO | \
RICHACE_SPECIAL_WHO)

#define RICHACE_INHERITANCE_FLAGS ( \
@@ -310,14 +314,28 @@ richace_is_deny(const struct richace *ace)
* richace_is_same_identifier - are both identifiers the same?
*/
static inline bool
-richace_is_same_identifier(const struct richace *a, const struct richace *b)
+richace_is_same_identifier(const struct richacl *acl,
+ const struct richace *ace1,
+ const struct richace *ace2)
{
- return !((a->e_flags ^ b->e_flags) &
- (RICHACE_SPECIAL_WHO | RICHACE_IDENTIFIER_GROUP)) &&
- !memcmp(&a->e_id, &b->e_id, sizeof(a->e_id));
+ const char *unmapped = (char *)(acl->a_entries + acl->a_count);
+
+ return !((ace1->e_flags ^ ace2->e_flags) &
+ (RICHACE_SPECIAL_WHO |
+ RICHACE_IDENTIFIER_GROUP |
+ RICHACE_UNMAPPED_WHO)) &&
+ ((ace1->e_flags & RICHACE_UNMAPPED_WHO) ?
+ !strcmp(unmapped + ace1->e_id.offs,
+ unmapped + ace2->e_id.offs) :
+ !memcmp(&ace1->e_id, &ace2->e_id, sizeof(ace1->e_id)));
+}
+
+extern struct richacl *__richacl_alloc(unsigned int, size_t, gfp_t);
+static inline struct richacl *richacl_alloc(unsigned int count, gfp_t gfp)
+{
+ return __richacl_alloc(count, 0, gfp);
}

-extern struct richacl *richacl_alloc(int, gfp_t);
extern struct richacl *richacl_clone(const struct richacl *, gfp_t);
extern void richace_copy(struct richace *, const struct richace *);
extern int richacl_masks_to_mode(const struct richacl *);
@@ -327,6 +345,11 @@ extern void richacl_compute_max_masks(struct richacl *);
extern struct richacl *richacl_chmod(struct richacl *, mode_t);
extern int richacl_equiv_mode(const struct richacl *, mode_t *);
extern struct richacl *richacl_inherit(const struct richacl *, int);
+extern int richacl_add_unmapped_identifier(struct richacl **, struct richace **,
+ const char *, unsigned int, gfp_t);
+extern const char *richace_unmapped_identifier(const struct richace *,
+ const struct richacl *);
+extern bool richacl_has_unmapped_identifiers(struct richacl *);

/* richacl_inode.c */
extern int richacl_permission(struct inode *, const struct richacl *, int);
--
2.4.3


2015-09-27 22:12:49

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 32/41] nfsd: Add support for the MAY_CREATE_{FILE,DIR} permissions

For local file systems, the vfs performs the necessary permission checks
for operations like creating files and directories. NFSd duplicates
several of those checks. The vfs checks have been extended to check for
additional permissions like MAY_CREATE_FILE and MY_CREATE_DIR; the nfsd
checks currently lack those extensions.

Ideally, all duplicate checks should be removed; for now, just fix the
duplicate checks instead though.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Acked-by: J. Bruce Fields <[email protected]>
---
fs/nfsd/nfs4proc.c | 5 +++--
fs/nfsd/nfsfh.c | 8 ++++----
fs/nfsd/vfs.c | 28 ++++++++++++++++++++--------
fs/nfsd/vfs.h | 17 +++++++++--------
4 files changed, 36 insertions(+), 22 deletions(-)

diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index a053e78..8d476ff 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -599,14 +599,15 @@ static __be32
nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
struct nfsd4_create *create)
{
+ int access = create->cr_type == NF4DIR ?
+ NFSD_MAY_CREATE_DIR : NFSD_MAY_CREATE_FILE;
struct svc_fh resfh;
__be32 status;
dev_t rdev;

fh_init(&resfh, NFS4_FHSIZE);

- status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR,
- NFSD_MAY_CREATE);
+ status = fh_verify(rqstp, &cstate->current_fh, S_IFDIR, access);
if (status)
return status;

diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c
index 350041a..7159316 100644
--- a/fs/nfsd/nfsfh.c
+++ b/fs/nfsd/nfsfh.c
@@ -319,10 +319,10 @@ fh_verify(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, int access)
/*
* We still have to do all these permission checks, even when
* fh_dentry is already set:
- * - fh_verify may be called multiple times with different
- * "access" arguments (e.g. nfsd_proc_create calls
- * fh_verify(...,NFSD_MAY_EXEC) first, then later (in
- * nfsd_create) calls fh_verify(...,NFSD_MAY_CREATE).
+ * - fh_verify may be called multiple times with different
+ * "access" arguments (e.g. nfsd_proc_create calls
+ * fh_verify(...,NFSD_MAY_EXEC) first, then later (in
+ * nfsd_create) calls fh_verify(...,NFSD_MAY_CREATE_FILE).
* - in the NFSv4 case, the filehandle may have been filled
* in by fh_compose, and given a dentry, but further
* compound operations performed with that filehandle
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 45c0497..fb35775 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1128,6 +1128,8 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
__be32 err;
__be32 err2;
int host_err;
+ int access = (type == S_IFDIR) ?
+ NFSD_MAY_CREATE_DIR : NFSD_MAY_CREATE_FILE;

err = nfserr_perm;
if (!flen)
@@ -1136,7 +1138,7 @@ nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,
if (isdotent(fname, flen))
goto out;

- err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE);
+ err = fh_verify(rqstp, fhp, S_IFDIR, access);
if (err)
goto out;

@@ -1301,7 +1303,7 @@ do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp,

/* If file doesn't exist, check for permissions to create one */
if (d_really_is_negative(dchild)) {
- err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE);
+ err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE_FILE);
if (err)
goto out;
}
@@ -1485,7 +1487,7 @@ nfsd_symlink(struct svc_rqst *rqstp, struct svc_fh *fhp,
if (isdotent(fname, flen))
goto out;

- err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE);
+ err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE_FILE);
if (err)
goto out;

@@ -1532,7 +1534,7 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp,
__be32 err;
int host_err;

- err = fh_verify(rqstp, ffhp, S_IFDIR, NFSD_MAY_CREATE);
+ err = fh_verify(rqstp, ffhp, S_IFDIR, NFSD_MAY_CREATE_FILE);
if (err)
goto out;
err = fh_verify(rqstp, tfhp, 0, NFSD_MAY_NOP);
@@ -1604,11 +1606,12 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
struct inode *fdir, *tdir;
__be32 err;
int host_err;
+ int access;

err = fh_verify(rqstp, ffhp, S_IFDIR, NFSD_MAY_REMOVE);
if (err)
goto out;
- err = fh_verify(rqstp, tfhp, S_IFDIR, NFSD_MAY_CREATE);
+ err = fh_verify(rqstp, tfhp, S_IFDIR, NFSD_MAY_NOP);
if (err)
goto out;

@@ -1647,6 +1650,13 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
if (odentry == trap)
goto out_dput_old;

+ host_err = 0;
+ access = S_ISDIR(d_inode(odentry)->i_mode) ?
+ NFSD_MAY_CREATE_DIR : NFSD_MAY_CREATE_FILE;
+ err = fh_verify(rqstp, tfhp, S_IFDIR, access);
+ if (err)
+ goto out_dput_old;
+
ndentry = lookup_one_len(tname, tdentry, tlen);
host_err = PTR_ERR(ndentry);
if (IS_ERR(ndentry))
@@ -1672,7 +1682,8 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
out_dput_old:
dput(odentry);
out_nfserr:
- err = nfserrno(host_err);
+ if (host_err)
+ err = nfserrno(host_err);
/*
* We cannot rely on fh_unlock on the two filehandles,
* as that would do the wrong thing if the two directories
@@ -2005,8 +2016,9 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp,
uid_eq(inode->i_uid, current_fsuid()))
return 0;

- /* This assumes NFSD_MAY_{READ,WRITE,EXEC} == MAY_{READ,WRITE,EXEC} */
- err = inode_permission(inode, acc & (MAY_READ|MAY_WRITE|MAY_EXEC));
+ /* This assumes NFSD_MAY_{READ,WRITE,EXEC} == MAY_{READ,WRITE,EXEC}. */
+ err = inode_permission(inode, acc & (MAY_READ|MAY_WRITE|MAY_EXEC|
+ MAY_CREATE_DIR|MAY_CREATE_FILE));

/* Allow read access to binaries even when mode 111 */
if (err == -EACCES && S_ISREG(inode->i_mode) &&
diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h
index fee2451..c849ef2 100644
--- a/fs/nfsd/vfs.h
+++ b/fs/nfsd/vfs.h
@@ -19,18 +19,19 @@
#define NFSD_MAY_TRUNC 0x010
#define NFSD_MAY_LOCK 0x020
#define NFSD_MAY_MASK 0x03f
+#define NFSD_MAY_CREATE_FILE 0x103 /* == MAY_{EXEC|WRITE|CREATE_FILE} */
+#define NFSD_MAY_CREATE_DIR 0x203 /* == MAY_{EXEC|WRITE|CREATE_DIR} */

/* extra hints to permission and open routines: */
-#define NFSD_MAY_OWNER_OVERRIDE 0x040
-#define NFSD_MAY_LOCAL_ACCESS 0x080 /* for device special files */
-#define NFSD_MAY_BYPASS_GSS_ON_ROOT 0x100
-#define NFSD_MAY_NOT_BREAK_LEASE 0x200
-#define NFSD_MAY_BYPASS_GSS 0x400
-#define NFSD_MAY_READ_IF_EXEC 0x800
+#define NFSD_MAY_OWNER_OVERRIDE 0x04000
+#define NFSD_MAY_LOCAL_ACCESS 0x08000 /* for device special files */
+#define NFSD_MAY_BYPASS_GSS_ON_ROOT 0x10000
+#define NFSD_MAY_NOT_BREAK_LEASE 0x20000
+#define NFSD_MAY_BYPASS_GSS 0x40000
+#define NFSD_MAY_READ_IF_EXEC 0x80000

-#define NFSD_MAY_64BIT_COOKIE 0x1000 /* 64 bit readdir cookies for >= NFSv3 */
+#define NFSD_MAY_64BIT_COOKIE 0x100000 /* 64 bit readdir cookies for >= NFSv3 */

-#define NFSD_MAY_CREATE (NFSD_MAY_EXEC|NFSD_MAY_WRITE)
#define NFSD_MAY_REMOVE (NFSD_MAY_EXEC|NFSD_MAY_WRITE|NFSD_MAY_TRUNC)

/*
--
2.4.3


2015-09-27 22:12:43

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 31/41] nfsd: Add support for the v4.1 dacl attribute

Richacls support the Automatic Inheritance permission propagation
mechanism as specified in NFSv4.1. Over NFS, this requires support for
the dacl attribute: compared to the acl attribute, the dacl attribute
has an additional flags field which indicates when Automatic Inheritance
is in use.

The server will only indicate dacl attribute support in protocol version
4.1 and later, on file systems with richacl support.

This commit also adds support for the NFSv4.1 NFS4_ACE_WRITE_RETENTION
and NFS4_ACE_WRITE_RETENTION_HOLD ACL permissions.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Acked-by: J. Bruce Fields <[email protected]>
---
fs/nfsd/nfs4proc.c | 3 +-
fs/nfsd/nfs4xdr.c | 219 ++++++++++++++++++++++++++++++----------------
fs/nfsd/nfsd.h | 6 +-
include/linux/nfs4.h | 1 +
include/uapi/linux/nfs4.h | 3 +-
5 files changed, 155 insertions(+), 77 deletions(-)

diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 1bcfda2..a053e78 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -1781,7 +1781,8 @@ static inline u32 nfsd4_getattr_rsize(struct svc_rqst *rqstp,
u32 bmap0 = bmap[0], bmap1 = bmap[1], bmap2 = bmap[2];
u32 ret = 0;

- if (bmap0 & FATTR4_WORD0_ACL)
+ if (bmap0 & FATTR4_WORD0_ACL ||
+ bmap1 & FATTR4_WORD1_DACL)
return svc_max_payload(rqstp);
if (bmap0 & FATTR4_WORD0_FS_LOCATIONS)
return svc_max_payload(rqstp);
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index 682a7d8..33d028c 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -301,6 +301,68 @@ nfsd4_decode_bitmap(struct nfsd4_compoundargs *argp, u32 *bmval)
DECODE_TAIL;
}

+static unsigned int
+nfsd4_ace_mask(int minorversion)
+{
+ return minorversion == 0 ? NFS40_ACE_MASK_ALL : NFS4_ACE_MASK_ALL;
+}
+
+static __be32
+nfsd4_decode_acl_entries(struct nfsd4_compoundargs *argp, struct richacl **acl,
+ unsigned short flags_mask, unsigned int ace_mask,
+ int *plen)
+{
+ struct richace *ace;
+ u32 dummy32;
+ char *buf;
+ int len = 0;
+
+ DECODE_HEAD;
+
+ flags_mask &= RICHACE_VALID_FLAGS & ~RICHACE_SPECIAL_WHO;
+
+ READ_BUF(4); len += 4;
+ dummy32 = be32_to_cpup(p++);
+
+ if (dummy32 > NFSD4_ACL_MAX)
+ return nfserr_fbig;
+
+ *acl = svcxdr_alloc_richacl(argp, dummy32);
+ if (*acl == NULL)
+ return nfserr_jukebox;
+
+ richacl_for_each_entry(ace, *acl) {
+ READ_BUF(16); len += 16;
+
+ dummy32 = be32_to_cpup(p++);
+ if (dummy32 > RICHACE_ACCESS_DENIED_ACE_TYPE)
+ return nfserr_inval;
+ ace->e_type = dummy32;
+
+ dummy32 = be32_to_cpup(p++);
+ if (dummy32 & ~flags_mask)
+ return nfserr_inval;
+ ace->e_flags = dummy32;
+
+ dummy32 = be32_to_cpup(p++);
+ if (dummy32 & ~ace_mask)
+ return nfserr_inval;
+ ace->e_mask = dummy32;
+
+ dummy32 = be32_to_cpup(p++);
+ READ_BUF(dummy32);
+ len += XDR_QUADLEN(dummy32) << 2;
+ READMEM(buf, dummy32);
+ status = nfsd4_decode_ace_who(ace, argp->rqstp,
+ buf, dummy32);
+ if (status)
+ return status;
+ }
+ *plen += len;
+
+ DECODE_TAIL;
+}
+
static __be32
nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,
struct iattr *iattr, struct richacl **acl,
@@ -312,6 +374,7 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,

DECODE_HEAD;
iattr->ia_valid = 0;
+ *acl = NULL;
if ((status = nfsd4_decode_bitmap(argp, bmval)))
return status;

@@ -325,50 +388,18 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,
iattr->ia_valid |= ATTR_SIZE;
}
if (bmval[0] & FATTR4_WORD0_ACL) {
- u32 nace;
- struct richace *ace;
-
- READ_BUF(4); len += 4;
- nace = be32_to_cpup(p++);
-
- if (nace > NFSD4_ACL_MAX)
- return nfserr_fbig;
+ if (bmval[1] & FATTR4_WORD1_DACL)
+ return nfserr_inval;

- *acl = svcxdr_alloc_richacl(argp, nace);
- if (*acl == NULL)
+ status = nfsd4_decode_acl_entries(argp, acl,
+ ~NFS4_ACE_INHERITED_ACE,
+ nfsd4_ace_mask(argp->minorversion),
+ &len);
+ if (status)
+ return status;
+ else if (*acl == NULL)
return nfserr_jukebox;
-
- richacl_for_each_entry(ace, *acl) {
- READ_BUF(16); len += 16;
-
- dummy32 = be32_to_cpup(p++);
- if (dummy32 > RICHACE_ACCESS_DENIED_ACE_TYPE)
- return nfserr_inval;
- ace->e_type = dummy32;
-
- dummy32 = be32_to_cpup(p++);
- if (dummy32 & (~RICHACE_VALID_FLAGS |
- RICHACE_INHERITED_ACE |
- RICHACE_SPECIAL_WHO))
- return nfserr_inval;
- ace->e_flags = dummy32;
-
- dummy32 = be32_to_cpup(p++);
- if (dummy32 & ~NFS4_ACE_MASK_ALL)
- return nfserr_inval;
- ace->e_mask = dummy32;
-
- dummy32 = be32_to_cpup(p++);
- READ_BUF(dummy32);
- len += XDR_QUADLEN(dummy32) << 2;
- READMEM(buf, dummy32);
- status = nfsd4_decode_ace_who(ace, argp->rqstp,
- buf, dummy32);
- if (status)
- return status;
- }
- } else
- *acl = NULL;
+ }
if (bmval[1] & FATTR4_WORD1_MODE) {
READ_BUF(4);
len += 4;
@@ -436,6 +467,22 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,
goto xdr_error;
}
}
+ if (bmval[1] & FATTR4_WORD1_DACL) {
+ READ_BUF(4);
+ len += 4;
+ dummy32 = be32_to_cpup(p++);
+ if (dummy32 & (~RICHACL_VALID_FLAGS | RICHACL_MASKED))
+ return nfserr_inval;
+ status = nfsd4_decode_acl_entries(argp, acl,
+ ~0,
+ nfsd4_ace_mask(argp->minorversion),
+ &len);
+ if (status)
+ return status;
+ else if (*acl == NULL)
+ return nfserr_jukebox;
+ (*acl)->a_flags = dummy32;
+ }

label->len = 0;
#ifdef CONFIG_NFSD_V4_SECURITY_LABEL
@@ -2272,6 +2319,42 @@ out_resource:
return nfserr_resource;
}

+static __be32 nfsd4_encode_acl_entries(struct xdr_stream *xdr,
+ struct richacl *acl, struct svc_rqst *rqstp,
+ unsigned short flags_mask, unsigned int ace_mask)
+{
+ __be32 *p;
+
+ flags_mask &= ~RICHACE_SPECIAL_WHO;
+
+ p = xdr_reserve_space(xdr, 4);
+ if (!p)
+ return nfserr_resource;
+
+ if (acl == NULL) {
+ *p++ = cpu_to_be32(0);
+ } else {
+ struct richace *ace;
+
+ *p++ = cpu_to_be32(acl->a_count);
+
+ richacl_for_each_entry(ace, acl) {
+ __be32 status;
+
+ p = xdr_reserve_space(xdr, 4*3);
+ if (!p)
+ return nfserr_resource;
+ *p++ = cpu_to_be32(ace->e_type);
+ *p++ = cpu_to_be32(ace->e_flags & flags_mask);
+ *p++ = cpu_to_be32(ace->e_mask & ace_mask);
+ status = nfsd4_encode_ace_who(xdr, rqstp, ace);
+ if (status)
+ return status;
+ }
+ }
+ return 0;
+}
+
/*
* Note: @fhp can be NULL; in this case, we might have to compose the filehandle
* ourselves.
@@ -2342,15 +2425,16 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
goto out;
fhp = tempfh;
}
- if (bmval0 & FATTR4_WORD0_ACL) {
+ if ((bmval0 & FATTR4_WORD0_ACL) || (bmval1 & FATTR4_WORD1_DACL)) {
acl = nfsd4_get_acl(rqstp, dentry);
if (IS_ERR(acl)) {
err = PTR_ERR(acl);
acl = NULL;
}
- if (err == -EOPNOTSUPP)
+ if (err == -EOPNOTSUPP) {
bmval0 &= ~FATTR4_WORD0_ACL;
- else if (err == -EINVAL) {
+ bmval1 &= ~FATTR4_WORD1_DACL;
+ } else if (err == -EINVAL) {
status = nfserr_attrnotsupp;
goto out;
} else if (err != 0)
@@ -2389,6 +2473,8 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,

if (!IS_ACL(d_inode(dentry)))
word0 &= ~FATTR4_WORD0_ACL;
+ if (!IS_RICHACL(d_inode(dentry)))
+ word1 &= ~FATTR4_WORD1_DACL;
if (!contextsupport)
word2 &= ~FATTR4_WORD2_SECURITY_LABEL;
if (!word2) {
@@ -2502,35 +2588,12 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
*p++ = cpu_to_be32(rdattr_err);
}
if (bmval0 & FATTR4_WORD0_ACL) {
- struct richace *ace;
-
- if (acl == NULL) {
- p = xdr_reserve_space(xdr, 4);
- if (!p)
- goto out_resource;
-
- *p++ = cpu_to_be32(0);
- goto out_acl;
- }
- p = xdr_reserve_space(xdr, 4);
- if (!p)
- goto out_resource;
- *p++ = cpu_to_be32(acl->a_count);
-
- richacl_for_each_entry(ace, acl) {
- p = xdr_reserve_space(xdr, 4*3);
- if (!p)
- goto out_resource;
- *p++ = cpu_to_be32(ace->e_type);
- *p++ = cpu_to_be32(ace->e_flags &
- ~(RICHACE_SPECIAL_WHO | RICHACE_INHERITED_ACE));
- *p++ = cpu_to_be32(ace->e_mask & NFS4_ACE_MASK_ALL);
- status = nfsd4_encode_ace_who(xdr, rqstp, ace);
- if (status)
- goto out;
- }
+ status = nfsd4_encode_acl_entries(xdr, acl, rqstp,
+ ~NFS4_ACE_INHERITED_ACE,
+ nfsd4_ace_mask(minorversion));
+ if (status)
+ goto out;
}
-out_acl:
if (bmval0 & FATTR4_WORD0_ACLSUPPORT) {
p = xdr_reserve_space(xdr, 4);
if (!p)
@@ -2746,6 +2809,16 @@ out_acl:
}
p = xdr_encode_hyper(p, ino);
}
+ if (bmval1 & FATTR4_WORD1_DACL) {
+ p = xdr_reserve_space(xdr, 4);
+ if (!p)
+ goto out_resource;
+ *p++ = cpu_to_be32(acl->a_flags);
+ status = nfsd4_encode_acl_entries(xdr, acl, rqstp,
+ ~0, nfsd4_ace_mask(minorversion));
+ if (status)
+ goto out;
+ }
#ifdef CONFIG_NFSD_PNFS
if (bmval1 & FATTR4_WORD1_FS_LAYOUT_TYPES) {
status = nfsd4_encode_layout_type(xdr, exp->ex_layout_type);
diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h
index cf98052..cb5c3ed 100644
--- a/fs/nfsd/nfsd.h
+++ b/fs/nfsd/nfsd.h
@@ -339,7 +339,8 @@ void nfsd_lockd_shutdown(void);
NFSD4_SUPPORTED_ATTRS_WORD0

#define NFSD4_1_SUPPORTED_ATTRS_WORD1 \
- (NFSD4_SUPPORTED_ATTRS_WORD1 | PNFSD_SUPPORTED_ATTRS_WORD1)
+ (NFSD4_SUPPORTED_ATTRS_WORD1 | PNFSD_SUPPORTED_ATTRS_WORD1 | \
+ FATTR4_WORD1_DACL)

#define NFSD4_1_SUPPORTED_ATTRS_WORD2 \
(NFSD4_SUPPORTED_ATTRS_WORD2 | PNFSD_SUPPORTED_ATTRS_WORD2 | \
@@ -386,7 +387,8 @@ static inline u32 nfsd_suppattrs2(u32 minorversion)
(FATTR4_WORD0_SIZE | FATTR4_WORD0_ACL)
#define NFSD_WRITEABLE_ATTRS_WORD1 \
(FATTR4_WORD1_MODE | FATTR4_WORD1_OWNER | FATTR4_WORD1_OWNER_GROUP \
- | FATTR4_WORD1_TIME_ACCESS_SET | FATTR4_WORD1_TIME_MODIFY_SET)
+ | FATTR4_WORD1_TIME_ACCESS_SET | FATTR4_WORD1_TIME_MODIFY_SET \
+ | FATTR4_WORD1_DACL)
#ifdef CONFIG_NFSD_V4_SECURITY_LABEL
#define NFSD_WRITEABLE_ATTRS_WORD2 FATTR4_WORD2_SECURITY_LABEL
#else
diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h
index 1422fc6..682ced3 100644
--- a/include/linux/nfs4.h
+++ b/include/linux/nfs4.h
@@ -394,6 +394,7 @@ enum lock_type4 {
#define FATTR4_WORD1_TIME_MODIFY (1UL << 21)
#define FATTR4_WORD1_TIME_MODIFY_SET (1UL << 22)
#define FATTR4_WORD1_MOUNTED_ON_FILEID (1UL << 23)
+#define FATTR4_WORD1_DACL (1UL << 26)
#define FATTR4_WORD1_FS_LAYOUT_TYPES (1UL << 30)
#define FATTR4_WORD2_LAYOUT_TYPES (1UL << 0)
#define FATTR4_WORD2_LAYOUT_BLKSIZE (1UL << 1)
diff --git a/include/uapi/linux/nfs4.h b/include/uapi/linux/nfs4.h
index 2b871e0..b850ffd 100644
--- a/include/uapi/linux/nfs4.h
+++ b/include/uapi/linux/nfs4.h
@@ -121,7 +121,8 @@
#define NFS4_ACE_GENERIC_READ 0x00120081
#define NFS4_ACE_GENERIC_WRITE 0x00160106
#define NFS4_ACE_GENERIC_EXECUTE 0x001200A0
-#define NFS4_ACE_MASK_ALL 0x001F01FF
+#define NFS40_ACE_MASK_ALL 0x001F01FF
+#define NFS4_ACE_MASK_ALL 0x001F07FF

#define EXCHGID4_FLAG_SUPP_MOVED_REFER 0x00000001
#define EXCHGID4_FLAG_SUPP_MOVED_MIGR 0x00000002
--
2.4.3


2015-09-27 22:12:38

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 30/41] nfsd: Add richacl support

On file systems with richacls enabled, get and set richacls directly
instead of converting from / to posix acls.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Acked-by: J. Bruce Fields <[email protected]>
---
fs/nfsd/acl.h | 3 +-
fs/nfsd/nfs4acl.c | 124 ++++++++++++++++++++++++++++++++++++++---------------
fs/nfsd/nfs4proc.c | 2 +-
fs/nfsd/nfs4xdr.c | 34 +++++++++++----
4 files changed, 117 insertions(+), 46 deletions(-)

diff --git a/fs/nfsd/acl.h b/fs/nfsd/acl.h
index 1c5deb5..d73c664 100644
--- a/fs/nfsd/acl.h
+++ b/fs/nfsd/acl.h
@@ -53,8 +53,7 @@ __be32 nfsd4_decode_ace_who(struct richace *ace, struct svc_rqst *rqstp,
__be32 nfsd4_encode_ace_who(struct xdr_stream *xdr, struct svc_rqst *rqstp,
struct richace *ace);

-int nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry,
- struct richacl **acl);
+struct richacl *nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry);
__be32 nfsd4_set_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
struct richacl *acl);

diff --git a/fs/nfsd/nfs4acl.c b/fs/nfsd/nfs4acl.c
index 6d3bb72..f017a76 100644
--- a/fs/nfsd/nfs4acl.c
+++ b/fs/nfsd/nfs4acl.c
@@ -40,6 +40,8 @@
#include <linux/nfs_fs.h>
#include <linux/richacl_compat.h>
#include <linux/nfs4acl.h>
+#include <linux/xattr.h>
+#include <linux/richacl_xattr.h>

#include "nfsfh.h"
#include "nfsd.h"
@@ -129,32 +131,28 @@ static short ace2type(struct richace *);
static void _posix_to_richacl_one(struct posix_acl *, struct richacl_alloc *,
unsigned int);

-int
-nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry,
- struct richacl **acl)
+static struct richacl *
+nfsd4_get_posix_acl(struct svc_rqst *rqstp, struct dentry *dentry)
{
struct inode *inode = d_inode(dentry);
- int error = 0;
struct posix_acl *pacl = NULL, *dpacl = NULL;
struct richacl_alloc alloc;
unsigned int flags = 0;
int count;

pacl = get_acl(inode, ACL_TYPE_ACCESS);
- if (!pacl)
- pacl = posix_acl_from_mode(inode->i_mode, GFP_KERNEL);
-
- if (IS_ERR(pacl))
- return PTR_ERR(pacl);
+ if (IS_ERR_OR_NULL(pacl))
+ return (void *)pacl;

- /* allocate for worst case: one (deny, allow) pair each: */
+ /* Allocate for worst case: one (deny, allow) pair each. The resulting
+ acl will be released shortly and won't be cached. */
count = 2 * pacl->a_count;

if (S_ISDIR(inode->i_mode)) {
flags = FLAG_DIRECTORY;
dpacl = get_acl(inode, ACL_TYPE_DEFAULT);
if (IS_ERR(dpacl)) {
- error = PTR_ERR(dpacl);
+ alloc.acl = (void *)dpacl;
goto rel_pacl;
}

@@ -163,7 +161,7 @@ nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry,
}

if (!richacl_prepare(&alloc, count)) {
- error = -ENOMEM;
+ alloc.acl = ERR_PTR(-ENOMEM);
goto out;
}

@@ -172,13 +170,37 @@ nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry,
if (dpacl)
_posix_to_richacl_one(dpacl, &alloc, flags | FLAG_DEFAULT_ACL);

- *acl = alloc.acl;
-
out:
posix_acl_release(dpacl);
rel_pacl:
posix_acl_release(pacl);
- return error;
+ return alloc.acl;
+}
+
+struct richacl *
+nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry)
+{
+ struct inode *inode = d_inode(dentry);
+ struct richacl *acl;
+ int error;
+
+ if (IS_RICHACL(inode))
+ acl = get_richacl(inode);
+ else
+ acl = nfsd4_get_posix_acl(rqstp, dentry);
+ if (IS_ERR(acl))
+ return acl;
+ else if (acl == NULL) {
+ acl = richacl_from_mode(inode->i_mode);
+ if (acl == NULL)
+ acl = ERR_PTR(-ENOMEM);
+ }
+ error = richacl_apply_masks(&acl, inode->i_uid);
+ if (error) {
+ richacl_put(acl);
+ acl = ERR_PTR(error);
+ }
+ return acl;
}

struct posix_acl_summary {
@@ -744,56 +766,88 @@ out_estate:
return ret;
}

-__be32
-nfsd4_set_acl(struct svc_rqst *rqstp, struct svc_fh *fhp, struct richacl *acl)
+static int
+nfsd4_set_posix_acl(struct svc_rqst *rqstp, struct dentry *dentry,
+ struct richacl *acl)
{
- __be32 error;
int host_error;
- struct dentry *dentry;
- struct inode *inode;
+ struct inode *inode = d_inode(dentry);
struct posix_acl *pacl = NULL, *dpacl = NULL;
unsigned int flags = 0;

- /* Get inode */
- error = fh_verify(rqstp, fhp, 0, NFSD_MAY_SATTR);
- if (error)
- return error;
-
- dentry = fhp->fh_dentry;
- inode = d_inode(dentry);
-
if (!inode->i_op->set_acl || !IS_POSIXACL(inode))
- return nfserr_attrnotsupp;
+ return -EOPNOTSUPP;

if (S_ISDIR(inode->i_mode))
flags = FLAG_DIRECTORY;

host_error = nfs4_richacl_to_posix(acl, &pacl, &dpacl, flags);
if (host_error == -EINVAL)
- return nfserr_attrnotsupp;
+ return -EOPNOTSUPP;
if (host_error < 0)
- goto out_nfserr;
+ return host_error;

host_error = inode->i_op->set_acl(inode, pacl, ACL_TYPE_ACCESS);
if (host_error < 0)
goto out_release;

- if (S_ISDIR(inode->i_mode)) {
+ if (S_ISDIR(inode->i_mode))
host_error = inode->i_op->set_acl(inode, dpacl,
ACL_TYPE_DEFAULT);
- }

out_release:
posix_acl_release(pacl);
posix_acl_release(dpacl);
-out_nfserr:
+ return host_error;
+}
+
+static int
+nfsd4_set_richacl(struct svc_rqst *rqstp, struct dentry *dentry,
+ struct richacl *acl)
+{
+ int host_error;
+ struct inode *inode = d_inode(dentry);
+ size_t size = richacl_xattr_size(acl);
+ char *buffer;
+
+ if (!inode->i_op->setxattr || !IS_RICHACL(inode))
+ return -EOPNOTSUPP;
+
+ richacl_compute_max_masks(acl);
+
+ buffer = kmalloc(size, GFP_KERNEL);
+ if (!buffer)
+ return -ENOMEM;
+ richacl_to_xattr(&init_user_ns, acl, buffer, size);
+ host_error = inode->i_op->setxattr(dentry, XATTR_NAME_RICHACL,
+ buffer, size, 0);
+ kfree(buffer);
+ return host_error;
+}
+
+__be32
+nfsd4_set_acl(struct svc_rqst *rqstp, struct svc_fh *fhp, struct richacl *acl)
+{
+ struct dentry *dentry;
+ int host_error;
+ __be32 error;
+
+ error = fh_verify(rqstp, fhp, 0, NFSD_MAY_SATTR);
+ if (error)
+ return error;
+ dentry = fhp->fh_dentry;
+
+ if (IS_RICHACL(d_inode(dentry)))
+ host_error = nfsd4_set_richacl(rqstp, dentry, acl);
+ else
+ host_error = nfsd4_set_posix_acl(rqstp, dentry, acl);
+
if (host_error == -EOPNOTSUPP)
return nfserr_attrnotsupp;
else
return nfserrno(host_error);
}

-
static short
ace2type(struct richace *ace)
{
diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 2430235..1bcfda2 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -110,7 +110,7 @@ check_attr_support(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
* in current environment or not.
*/
if (bmval[0] & FATTR4_WORD0_ACL) {
- if (!IS_POSIXACL(d_inode(dentry)))
+ if (!IS_ACL(d_inode(dentry)))
return nfserr_attrnotsupp;
}

diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index 8603f40..682a7d8 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -340,11 +340,24 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,

richacl_for_each_entry(ace, *acl) {
READ_BUF(16); len += 16;
- ace->e_type = be32_to_cpup(p++);
- ace->e_flags = be32_to_cpup(p++);
- ace->e_mask = be32_to_cpup(p++);
- if (ace->e_flags & RICHACE_SPECIAL_WHO)
+
+ dummy32 = be32_to_cpup(p++);
+ if (dummy32 > RICHACE_ACCESS_DENIED_ACE_TYPE)
+ return nfserr_inval;
+ ace->e_type = dummy32;
+
+ dummy32 = be32_to_cpup(p++);
+ if (dummy32 & (~RICHACE_VALID_FLAGS |
+ RICHACE_INHERITED_ACE |
+ RICHACE_SPECIAL_WHO))
return nfserr_inval;
+ ace->e_flags = dummy32;
+
+ dummy32 = be32_to_cpup(p++);
+ if (dummy32 & ~NFS4_ACE_MASK_ALL)
+ return nfserr_inval;
+ ace->e_mask = dummy32;
+
dummy32 = be32_to_cpup(p++);
READ_BUF(dummy32);
len += XDR_QUADLEN(dummy32) << 2;
@@ -2330,7 +2343,11 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
fhp = tempfh;
}
if (bmval0 & FATTR4_WORD0_ACL) {
- err = nfsd4_get_acl(rqstp, dentry, &acl);
+ acl = nfsd4_get_acl(rqstp, dentry);
+ if (IS_ERR(acl)) {
+ err = PTR_ERR(acl);
+ acl = NULL;
+ }
if (err == -EOPNOTSUPP)
bmval0 &= ~FATTR4_WORD0_ACL;
else if (err == -EINVAL) {
@@ -2370,7 +2387,7 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
u32 word1 = nfsd_suppattrs1(minorversion);
u32 word2 = nfsd_suppattrs2(minorversion);

- if (!IS_POSIXACL(dentry->d_inode))
+ if (!IS_ACL(d_inode(dentry)))
word0 &= ~FATTR4_WORD0_ACL;
if (!contextsupport)
word2 &= ~FATTR4_WORD2_SECURITY_LABEL;
@@ -2505,7 +2522,8 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
if (!p)
goto out_resource;
*p++ = cpu_to_be32(ace->e_type);
- *p++ = cpu_to_be32(ace->e_flags & ~RICHACE_SPECIAL_WHO);
+ *p++ = cpu_to_be32(ace->e_flags &
+ ~(RICHACE_SPECIAL_WHO | RICHACE_INHERITED_ACE));
*p++ = cpu_to_be32(ace->e_mask & NFS4_ACE_MASK_ALL);
status = nfsd4_encode_ace_who(xdr, rqstp, ace);
if (status)
@@ -2517,7 +2535,7 @@ out_acl:
p = xdr_reserve_space(xdr, 4);
if (!p)
goto out_resource;
- *p++ = cpu_to_be32(IS_POSIXACL(dentry->d_inode) ?
+ *p++ = cpu_to_be32(IS_ACL(d_inode(dentry)) ?
ACL4_SUPPORT_ALLOW_ACL|ACL4_SUPPORT_DENY_ACL : 0);
}
if (bmval0 & FATTR4_WORD0_CANSETTIME) {
--
2.4.3


2015-09-27 22:12:32

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 29/41] nfsd: Use richacls as internal acl representation

When converting from NFSv4 ACLs to POSIX ACLs, nfsd so far was using
struct nfs4_acl as its internal representation. This representation is a
subset of richacls, so get rid of struct nfs4_acl. Richacls even have a
more compact in-memory representation, so a few more ACL entries can
easily be supported.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Acked-by: J. Bruce Fields <[email protected]>
---
fs/Kconfig | 6 +
fs/nfs_common/Makefile | 1 +
fs/nfs_common/nfs4acl.c | 44 ++++++
fs/nfsd/Kconfig | 1 +
fs/nfsd/acl.h | 24 ++--
fs/nfsd/nfs4acl.c | 368 ++++++++++++++++++++++--------------------------
fs/nfsd/nfs4proc.c | 15 +-
fs/nfsd/nfs4xdr.c | 64 +++------
fs/nfsd/xdr4.h | 6 +-
include/linux/nfs4.h | 23 ---
include/linux/nfs4acl.h | 7 +
11 files changed, 274 insertions(+), 285 deletions(-)
create mode 100644 fs/nfs_common/nfs4acl.c
create mode 100644 include/linux/nfs4acl.h

diff --git a/fs/Kconfig b/fs/Kconfig
index bff2879..68bc3e1 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -265,6 +265,12 @@ config NFS_COMMON
depends on NFSD || NFS_FS || LOCKD
default y

+config NFS_RICHACL
+ bool
+ depends on NFSD_V4 || NFS_V4
+ select FS_RICHACL
+ default y
+
source "net/sunrpc/Kconfig"
source "fs/ceph/Kconfig"
source "fs/cifs/Kconfig"
diff --git a/fs/nfs_common/Makefile b/fs/nfs_common/Makefile
index d153ca3..e055139 100644
--- a/fs/nfs_common/Makefile
+++ b/fs/nfs_common/Makefile
@@ -4,5 +4,6 @@

obj-$(CONFIG_NFS_ACL_SUPPORT) += nfs_acl.o
nfs_acl-objs := nfsacl.o
+obj-$(CONFIG_NFS_RICHACL) += nfs4acl.o

obj-$(CONFIG_GRACE_PERIOD) += grace.o
diff --git a/fs/nfs_common/nfs4acl.c b/fs/nfs_common/nfs4acl.c
new file mode 100644
index 0000000..02df064
--- /dev/null
+++ b/fs/nfs_common/nfs4acl.c
@@ -0,0 +1,44 @@
+#include <linux/fs.h>
+#include <linux/richacl.h>
+#include <linux/nfs4acl.h>
+
+static struct special_id {
+ char *who;
+ int len;
+} special_who_map[] = {
+ [RICHACE_OWNER_SPECIAL_ID] = {
+ .who = "OWNER@",
+ .len = sizeof("OWNER@") - 1 },
+ [RICHACE_GROUP_SPECIAL_ID] = {
+ .who = "GROUP@",
+ .len = sizeof("GROUP@") - 1 },
+ [RICHACE_EVERYONE_SPECIAL_ID] = {
+ .who = "EVERYONE@",
+ .len = sizeof("EVERYONE@") - 1 }
+};
+
+int nfs4acl_who_to_special_id(const char *who, u32 len)
+{
+ int n;
+
+ for (n = 0; n < ARRAY_SIZE(special_who_map); n++) {
+ if (len == special_who_map[n].len &&
+ !memcmp(who, special_who_map[n].who, len))
+ return n;
+ }
+ return -1;
+}
+EXPORT_SYMBOL(nfs4acl_who_to_special_id);
+
+bool nfs4acl_special_id_to_who(unsigned int special_who,
+ const char **who, unsigned int *len)
+{
+ struct special_id *special = &special_who_map[special_who];
+
+ if (special_who > ARRAY_SIZE(special_who_map) || !special->len)
+ return false;
+ *who = special->who;
+ *len = special->len;
+ return true;
+}
+EXPORT_SYMBOL(nfs4acl_special_id_to_who);
diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig
index a0b77fc..811379a 100644
--- a/fs/nfsd/Kconfig
+++ b/fs/nfsd/Kconfig
@@ -70,6 +70,7 @@ config NFSD_V4
depends on NFSD && PROC_FS
select NFSD_V3
select FS_POSIX_ACL
+ select FS_RICHACL
select SUNRPC_GSS
select CRYPTO
select GRACE_PERIOD
diff --git a/fs/nfsd/acl.h b/fs/nfsd/acl.h
index 4cd7c69..1c5deb5 100644
--- a/fs/nfsd/acl.h
+++ b/fs/nfsd/acl.h
@@ -35,25 +35,27 @@
#ifndef LINUX_NFS4_ACL_H
#define LINUX_NFS4_ACL_H

-struct nfs4_acl;
+struct richacl;
+struct richace;
struct svc_fh;
struct svc_rqst;

/*
* Maximum ACL we'll accept from a client; chosen (somewhat
* arbitrarily) so that kmalloc'ing the ACL shouldn't require a
- * high-order allocation. This allows 204 ACEs on x86_64:
+ * high-order allocation. This allows 339 ACEs on x86_64:
*/
-#define NFS4_ACL_MAX ((PAGE_SIZE - sizeof(struct nfs4_acl)) \
- / sizeof(struct nfs4_ace))
+#define NFSD4_ACL_MAX ((PAGE_SIZE - sizeof(struct richacl)) \
+ / sizeof(struct richace))

-int nfs4_acl_bytes(int entries);
-int nfs4_acl_get_whotype(char *, u32);
-__be32 nfs4_acl_write_who(struct xdr_stream *xdr, int who);
+__be32 nfsd4_decode_ace_who(struct richace *ace, struct svc_rqst *rqstp,
+ char *who, u32 len);
+__be32 nfsd4_encode_ace_who(struct xdr_stream *xdr, struct svc_rqst *rqstp,
+ struct richace *ace);

-int nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry,
- struct nfs4_acl **acl);
-__be32 nfsd4_set_nfs4_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
- struct nfs4_acl *acl);
+int nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry,
+ struct richacl **acl);
+__be32 nfsd4_set_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
+ struct richacl *acl);

#endif /* LINUX_NFS4_ACL_H */
diff --git a/fs/nfsd/nfs4acl.c b/fs/nfsd/nfs4acl.c
index 6adabd6..6d3bb72 100644
--- a/fs/nfsd/nfs4acl.c
+++ b/fs/nfsd/nfs4acl.c
@@ -37,46 +37,50 @@
#include <linux/fs.h>
#include <linux/slab.h>
#include <linux/posix_acl.h>
+#include <linux/nfs_fs.h>
+#include <linux/richacl_compat.h>
+#include <linux/nfs4acl.h>

#include "nfsfh.h"
#include "nfsd.h"
+#include "idmap.h"
#include "acl.h"
#include "vfs.h"

-#define NFS4_ACL_TYPE_DEFAULT 0x01
-#define NFS4_ACL_DIR 0x02
-#define NFS4_ACL_OWNER 0x04
+#define FLAG_DEFAULT_ACL 0x01
+#define FLAG_DIRECTORY 0x02
+#define FLAG_OWNER 0x04

/* mode bit translations: */
-#define NFS4_READ_MODE (NFS4_ACE_READ_DATA)
-#define NFS4_WRITE_MODE (NFS4_ACE_WRITE_DATA | NFS4_ACE_APPEND_DATA)
-#define NFS4_EXECUTE_MODE NFS4_ACE_EXECUTE
-#define NFS4_ANYONE_MODE (NFS4_ACE_READ_ATTRIBUTES | NFS4_ACE_READ_ACL | NFS4_ACE_SYNCHRONIZE)
-#define NFS4_OWNER_MODE (NFS4_ACE_WRITE_ATTRIBUTES | NFS4_ACE_WRITE_ACL)
+#define RICHACE_READ_MODE (RICHACE_READ_DATA)
+#define RICHACE_WRITE_MODE (RICHACE_WRITE_DATA | RICHACE_APPEND_DATA)
+#define RICHACE_EXECUTE_MODE RICHACE_EXECUTE
+#define RICHACE_ANYONE_MODE (RICHACE_READ_ATTRIBUTES | RICHACE_READ_ACL | RICHACE_SYNCHRONIZE)
+#define RICHACE_OWNER_MODE (RICHACE_WRITE_ATTRIBUTES | RICHACE_WRITE_ACL)

/* flags used to simulate posix default ACLs */
-#define NFS4_INHERITANCE_FLAGS (NFS4_ACE_FILE_INHERIT_ACE \
- | NFS4_ACE_DIRECTORY_INHERIT_ACE)
-
-#define NFS4_SUPPORTED_FLAGS (NFS4_INHERITANCE_FLAGS \
- | NFS4_ACE_INHERIT_ONLY_ACE \
- | NFS4_ACE_IDENTIFIER_GROUP)
+#define RICHACE_SUPPORTED_FLAGS ( \
+ RICHACE_FILE_INHERIT_ACE | \
+ RICHACE_DIRECTORY_INHERIT_ACE | \
+ RICHACE_INHERIT_ONLY_ACE | \
+ RICHACE_IDENTIFIER_GROUP | \
+ RICHACE_SPECIAL_WHO)

static u32
mask_from_posix(unsigned short perm, unsigned int flags)
{
- int mask = NFS4_ANYONE_MODE;
+ int mask = RICHACE_ANYONE_MODE;

- if (flags & NFS4_ACL_OWNER)
- mask |= NFS4_OWNER_MODE;
+ if (flags & FLAG_OWNER)
+ mask |= RICHACE_OWNER_MODE;
if (perm & ACL_READ)
- mask |= NFS4_READ_MODE;
+ mask |= RICHACE_READ_MODE;
if (perm & ACL_WRITE)
- mask |= NFS4_WRITE_MODE;
- if ((perm & ACL_WRITE) && (flags & NFS4_ACL_DIR))
- mask |= NFS4_ACE_DELETE_CHILD;
+ mask |= RICHACE_WRITE_MODE;
+ if ((perm & ACL_WRITE) && (flags & FLAG_DIRECTORY))
+ mask |= RICHACE_DELETE_CHILD;
if (perm & ACL_EXECUTE)
- mask |= NFS4_EXECUTE_MODE;
+ mask |= RICHACE_EXECUTE_MODE;
return mask;
}

@@ -86,13 +90,13 @@ deny_mask_from_posix(unsigned short perm, u32 flags)
u32 mask = 0;

if (perm & ACL_READ)
- mask |= NFS4_READ_MODE;
+ mask |= RICHACE_READ_MODE;
if (perm & ACL_WRITE)
- mask |= NFS4_WRITE_MODE;
- if ((perm & ACL_WRITE) && (flags & NFS4_ACL_DIR))
- mask |= NFS4_ACE_DELETE_CHILD;
+ mask |= RICHACE_WRITE_MODE;
+ if ((perm & ACL_WRITE) && (flags & FLAG_DIRECTORY))
+ mask |= RICHACE_DELETE_CHILD;
if (perm & ACL_EXECUTE)
- mask |= NFS4_EXECUTE_MODE;
+ mask |= RICHACE_EXECUTE_MODE;
return mask;
}

@@ -108,32 +112,33 @@ deny_mask_from_posix(unsigned short perm, u32 flags)
static void
low_mode_from_nfs4(u32 perm, unsigned short *mode, unsigned int flags)
{
- u32 write_mode = NFS4_WRITE_MODE;
+ u32 write_mode = RICHACE_WRITE_MODE;

- if (flags & NFS4_ACL_DIR)
- write_mode |= NFS4_ACE_DELETE_CHILD;
+ if (flags & FLAG_DIRECTORY)
+ write_mode |= RICHACE_DELETE_CHILD;
*mode = 0;
- if ((perm & NFS4_READ_MODE) == NFS4_READ_MODE)
+ if ((perm & RICHACE_READ_MODE) == RICHACE_READ_MODE)
*mode |= ACL_READ;
if ((perm & write_mode) == write_mode)
*mode |= ACL_WRITE;
- if ((perm & NFS4_EXECUTE_MODE) == NFS4_EXECUTE_MODE)
+ if ((perm & RICHACE_EXECUTE_MODE) == RICHACE_EXECUTE_MODE)
*mode |= ACL_EXECUTE;
}

-static short ace2type(struct nfs4_ace *);
-static void _posix_to_nfsv4_one(struct posix_acl *, struct nfs4_acl *,
+static short ace2type(struct richace *);
+static void _posix_to_richacl_one(struct posix_acl *, struct richacl_alloc *,
unsigned int);

int
-nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry,
- struct nfs4_acl **acl)
+nfsd4_get_acl(struct svc_rqst *rqstp, struct dentry *dentry,
+ struct richacl **acl)
{
struct inode *inode = d_inode(dentry);
int error = 0;
struct posix_acl *pacl = NULL, *dpacl = NULL;
+ struct richacl_alloc alloc;
unsigned int flags = 0;
- int size = 0;
+ int count;

pacl = get_acl(inode, ACL_TYPE_ACCESS);
if (!pacl)
@@ -143,10 +148,10 @@ nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry,
return PTR_ERR(pacl);

/* allocate for worst case: one (deny, allow) pair each: */
- size += 2 * pacl->a_count;
+ count = 2 * pacl->a_count;

if (S_ISDIR(inode->i_mode)) {
- flags = NFS4_ACL_DIR;
+ flags = FLAG_DIRECTORY;
dpacl = get_acl(inode, ACL_TYPE_DEFAULT);
if (IS_ERR(dpacl)) {
error = PTR_ERR(dpacl);
@@ -154,20 +159,20 @@ nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry,
}

if (dpacl)
- size += 2 * dpacl->a_count;
+ count += 2 * dpacl->a_count;
}

- *acl = kmalloc(nfs4_acl_bytes(size), GFP_KERNEL);
- if (*acl == NULL) {
+ if (!richacl_prepare(&alloc, count)) {
error = -ENOMEM;
goto out;
}
- (*acl)->naces = 0;

- _posix_to_nfsv4_one(pacl, *acl, flags & ~NFS4_ACL_TYPE_DEFAULT);
+ _posix_to_richacl_one(pacl, &alloc, flags);

if (dpacl)
- _posix_to_nfsv4_one(dpacl, *acl, flags | NFS4_ACL_TYPE_DEFAULT);
+ _posix_to_richacl_one(dpacl, &alloc, flags | FLAG_DEFAULT_ACL);
+
+ *acl = alloc.acl;

out:
posix_acl_release(dpacl);
@@ -230,21 +235,22 @@ summarize_posix_acl(struct posix_acl *acl, struct posix_acl_summary *pas)

/* We assume the acl has been verified with posix_acl_valid. */
static void
-_posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl,
- unsigned int flags)
+_posix_to_richacl_one(struct posix_acl *pacl, struct richacl_alloc *alloc,
+ unsigned int flags)
{
struct posix_acl_entry *pa, *group_owner_entry;
- struct nfs4_ace *ace;
+ struct richace *ace;
struct posix_acl_summary pas;
unsigned short deny;
- int eflag = ((flags & NFS4_ACL_TYPE_DEFAULT) ?
- NFS4_INHERITANCE_FLAGS | NFS4_ACE_INHERIT_ONLY_ACE : 0);
+ int e_flags = ((flags & FLAG_DEFAULT_ACL) ?
+ (RICHACE_FILE_INHERIT_ACE |
+ RICHACE_DIRECTORY_INHERIT_ACE |
+ RICHACE_INHERIT_ONLY_ACE) : 0);

BUG_ON(pacl->a_count < 3);
summarize_posix_acl(pacl, &pas);

pa = pacl->a_entries;
- ace = acl->aces + acl->naces;

/* We could deny everything not granted by the owner: */
deny = ~pas.owner;
@@ -254,42 +260,35 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl,
*/
deny &= pas.users | pas.group | pas.groups | pas.other;
if (deny) {
- ace->type = NFS4_ACE_ACCESS_DENIED_ACE_TYPE;
- ace->flag = eflag;
- ace->access_mask = deny_mask_from_posix(deny, flags);
- ace->whotype = NFS4_ACL_WHO_OWNER;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags = e_flags | RICHACE_SPECIAL_WHO;
+ ace->e_mask = deny_mask_from_posix(deny, flags);
+ ace->e_id.special = RICHACE_OWNER_SPECIAL_ID;
}

- ace->type = NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE;
- ace->flag = eflag;
- ace->access_mask = mask_from_posix(pa->e_perm, flags | NFS4_ACL_OWNER);
- ace->whotype = NFS4_ACL_WHO_OWNER;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = e_flags | RICHACE_SPECIAL_WHO;
+ ace->e_mask = mask_from_posix(pa->e_perm, flags | FLAG_OWNER);
+ ace->e_id.special = RICHACE_OWNER_SPECIAL_ID;
pa++;

while (pa->e_tag == ACL_USER) {
deny = ~(pa->e_perm & pas.mask);
deny &= pas.groups | pas.group | pas.other;
if (deny) {
- ace->type = NFS4_ACE_ACCESS_DENIED_ACE_TYPE;
- ace->flag = eflag;
- ace->access_mask = deny_mask_from_posix(deny, flags);
- ace->whotype = NFS4_ACL_WHO_NAMED;
- ace->who_uid = pa->e_uid;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags = e_flags;
+ ace->e_mask = deny_mask_from_posix(deny, flags);
+ ace->e_id.uid = pa->e_uid;
}
- ace->type = NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE;
- ace->flag = eflag;
- ace->access_mask = mask_from_posix(pa->e_perm & pas.mask,
- flags);
- ace->whotype = NFS4_ACL_WHO_NAMED;
- ace->who_uid = pa->e_uid;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = e_flags;
+ ace->e_mask = mask_from_posix(pa->e_perm & pas.mask, flags);
+ ace->e_id.uid = pa->e_uid;
pa++;
}

@@ -300,23 +299,19 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl,

group_owner_entry = pa;

- ace->type = NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE;
- ace->flag = eflag;
- ace->access_mask = mask_from_posix(pas.group, flags);
- ace->whotype = NFS4_ACL_WHO_GROUP;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = e_flags | RICHACE_SPECIAL_WHO;
+ ace->e_mask = mask_from_posix(pas.group, flags);
+ ace->e_id.special = RICHACE_GROUP_SPECIAL_ID;
pa++;

while (pa->e_tag == ACL_GROUP) {
- ace->type = NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE;
- ace->flag = eflag | NFS4_ACE_IDENTIFIER_GROUP;
- ace->access_mask = mask_from_posix(pa->e_perm & pas.mask,
- flags);
- ace->whotype = NFS4_ACL_WHO_NAMED;
- ace->who_gid = pa->e_gid;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = e_flags | RICHACE_IDENTIFIER_GROUP;
+ ace->e_mask = mask_from_posix(pa->e_perm & pas.mask, flags);
+ ace->e_id.gid = pa->e_gid;
pa++;
}

@@ -326,12 +321,11 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl,

deny = ~pas.group & pas.other;
if (deny) {
- ace->type = NFS4_ACE_ACCESS_DENIED_ACE_TYPE;
- ace->flag = eflag;
- ace->access_mask = deny_mask_from_posix(deny, flags);
- ace->whotype = NFS4_ACL_WHO_GROUP;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags = e_flags | RICHACE_SPECIAL_WHO;
+ ace->e_mask = deny_mask_from_posix(deny, flags);
+ ace->e_id.special = RICHACE_GROUP_SPECIAL_ID;
}
pa++;

@@ -339,24 +333,22 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl,
deny = ~(pa->e_perm & pas.mask);
deny &= pas.other;
if (deny) {
- ace->type = NFS4_ACE_ACCESS_DENIED_ACE_TYPE;
- ace->flag = eflag | NFS4_ACE_IDENTIFIER_GROUP;
- ace->access_mask = deny_mask_from_posix(deny, flags);
- ace->whotype = NFS4_ACL_WHO_NAMED;
- ace->who_gid = pa->e_gid;
- ace++;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags = e_flags | RICHACE_IDENTIFIER_GROUP;
+ ace->e_mask = deny_mask_from_posix(deny, flags);
+ ace->e_id.gid = pa->e_gid;
}
pa++;
}

if (pa->e_tag == ACL_MASK)
pa++;
- ace->type = NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE;
- ace->flag = eflag;
- ace->access_mask = mask_from_posix(pa->e_perm, flags);
- ace->whotype = NFS4_ACL_WHO_EVERYONE;
- acl->naces++;
+ ace = richacl_append_entry(alloc);
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = e_flags | RICHACE_SPECIAL_WHO;
+ ace->e_mask = mask_from_posix(pa->e_perm, flags);
+ ace->e_id.special = RICHACE_EVERYONE_SPECIAL_ID;
}

static bool
@@ -500,7 +492,7 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags)
* and effective cases: when there are no inheritable ACEs,
* calls ->set_acl with a NULL ACL structure.
*/
- if (state->empty && (flags & NFS4_ACL_TYPE_DEFAULT))
+ if (state->empty && (flags & FLAG_DEFAULT_ACL))
return NULL;

/*
@@ -619,24 +611,24 @@ static void allow_bits_array(struct posix_ace_state_array *a, u32 mask)
}

static void process_one_v4_ace(struct posix_acl_state *state,
- struct nfs4_ace *ace)
+ struct richace *ace)
{
- u32 mask = ace->access_mask;
+ u32 mask = ace->e_mask;
int i;

state->empty = 0;

switch (ace2type(ace)) {
case ACL_USER_OBJ:
- if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) {
+ if (ace->e_type == RICHACE_ACCESS_ALLOWED_ACE_TYPE) {
allow_bits(&state->owner, mask);
} else {
deny_bits(&state->owner, mask);
}
break;
case ACL_USER:
- i = find_uid(state, ace->who_uid);
- if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) {
+ i = find_uid(state, ace->e_id.uid);
+ if (ace->e_type == RICHACE_ACCESS_ALLOWED_ACE_TYPE) {
allow_bits(&state->users->aces[i].perms, mask);
} else {
deny_bits(&state->users->aces[i].perms, mask);
@@ -645,7 +637,7 @@ static void process_one_v4_ace(struct posix_acl_state *state,
}
break;
case ACL_GROUP_OBJ:
- if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) {
+ if (ace->e_type == RICHACE_ACCESS_ALLOWED_ACE_TYPE) {
allow_bits(&state->group, mask);
} else {
deny_bits(&state->group, mask);
@@ -657,8 +649,8 @@ static void process_one_v4_ace(struct posix_acl_state *state,
}
break;
case ACL_GROUP:
- i = find_gid(state, ace->who_gid);
- if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) {
+ i = find_gid(state, ace->e_id.gid);
+ if (ace->e_type == RICHACE_ACCESS_ALLOWED_ACE_TYPE) {
allow_bits(&state->groups->aces[i].perms, mask);
} else {
deny_bits(&state->groups->aces[i].perms, mask);
@@ -671,7 +663,7 @@ static void process_one_v4_ace(struct posix_acl_state *state,
}
break;
case ACL_OTHER:
- if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) {
+ if (ace->e_type == RICHACE_ACCESS_ALLOWED_ACE_TYPE) {
allow_bits(&state->owner, mask);
allow_bits(&state->group, mask);
allow_bits(&state->other, mask);
@@ -689,32 +681,33 @@ static void process_one_v4_ace(struct posix_acl_state *state,
}
}

-static int nfs4_acl_nfsv4_to_posix(struct nfs4_acl *acl,
+static int nfs4_richacl_to_posix(struct richacl *acl,
struct posix_acl **pacl, struct posix_acl **dpacl,
unsigned int flags)
{
struct posix_acl_state effective_acl_state, default_acl_state;
- struct nfs4_ace *ace;
+ struct richace *ace;
int ret;

- ret = init_state(&effective_acl_state, acl->naces);
+ ret = init_state(&effective_acl_state, acl->a_count);
if (ret)
return ret;
- ret = init_state(&default_acl_state, acl->naces);
+ ret = init_state(&default_acl_state, acl->a_count);
if (ret)
goto out_estate;
ret = -EINVAL;
- for (ace = acl->aces; ace < acl->aces + acl->naces; ace++) {
- if (ace->type != NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE &&
- ace->type != NFS4_ACE_ACCESS_DENIED_ACE_TYPE)
+ richacl_for_each_entry(ace, acl) {
+ if (ace->e_type != RICHACE_ACCESS_ALLOWED_ACE_TYPE &&
+ ace->e_type != RICHACE_ACCESS_DENIED_ACE_TYPE)
goto out_dstate;
- if (ace->flag & ~NFS4_SUPPORTED_FLAGS)
+ if (ace->e_flags & ~RICHACE_SUPPORTED_FLAGS)
goto out_dstate;
- if ((ace->flag & NFS4_INHERITANCE_FLAGS) == 0) {
+ if ((ace->e_flags & (RICHACE_FILE_INHERIT_ACE |
+ RICHACE_DIRECTORY_INHERIT_ACE)) == 0) {
process_one_v4_ace(&effective_acl_state, ace);
continue;
}
- if (!(flags & NFS4_ACL_DIR))
+ if (!(flags & FLAG_DIRECTORY))
goto out_dstate;
/*
* Note that when only one of FILE_INHERIT or DIRECTORY_INHERIT
@@ -723,7 +716,7 @@ static int nfs4_acl_nfsv4_to_posix(struct nfs4_acl *acl,
*/
process_one_v4_ace(&default_acl_state, ace);

- if (!(ace->flag & NFS4_ACE_INHERIT_ONLY_ACE))
+ if (!(ace->e_flags & RICHACE_INHERIT_ONLY_ACE))
process_one_v4_ace(&effective_acl_state, ace);
}
*pacl = posix_state_to_acl(&effective_acl_state, flags);
@@ -733,7 +726,7 @@ static int nfs4_acl_nfsv4_to_posix(struct nfs4_acl *acl,
goto out_dstate;
}
*dpacl = posix_state_to_acl(&default_acl_state,
- flags | NFS4_ACL_TYPE_DEFAULT);
+ flags | FLAG_DEFAULT_ACL);
if (IS_ERR(*dpacl)) {
ret = PTR_ERR(*dpacl);
*dpacl = NULL;
@@ -752,8 +745,7 @@ out_estate:
}

__be32
-nfsd4_set_nfs4_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
- struct nfs4_acl *acl)
+nfsd4_set_acl(struct svc_rqst *rqstp, struct svc_fh *fhp, struct richacl *acl)
{
__be32 error;
int host_error;
@@ -774,9 +766,9 @@ nfsd4_set_nfs4_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
return nfserr_attrnotsupp;

if (S_ISDIR(inode->i_mode))
- flags = NFS4_ACL_DIR;
+ flags = FLAG_DIRECTORY;

- host_error = nfs4_acl_nfsv4_to_posix(acl, &pacl, &dpacl, flags);
+ host_error = nfs4_richacl_to_posix(acl, &pacl, &dpacl, flags);
if (host_error == -EINVAL)
return nfserr_attrnotsupp;
if (host_error < 0)
@@ -803,82 +795,62 @@ out_nfserr:


static short
-ace2type(struct nfs4_ace *ace)
+ace2type(struct richace *ace)
{
- switch (ace->whotype) {
- case NFS4_ACL_WHO_NAMED:
- return (ace->flag & NFS4_ACE_IDENTIFIER_GROUP ?
- ACL_GROUP : ACL_USER);
- case NFS4_ACL_WHO_OWNER:
+ if (ace->e_flags & RICHACE_SPECIAL_WHO) {
+ switch (ace->e_id.special) {
+ case RICHACE_OWNER_SPECIAL_ID:
return ACL_USER_OBJ;
- case NFS4_ACL_WHO_GROUP:
+ case RICHACE_GROUP_SPECIAL_ID:
return ACL_GROUP_OBJ;
- case NFS4_ACL_WHO_EVERYONE:
+ case RICHACE_EVERYONE_SPECIAL_ID:
return ACL_OTHER;
+ default:
+ BUG();
+ }
}
- BUG();
- return -1;
-}
-
-/*
- * return the size of the struct nfs4_acl required to represent an acl
- * with @entries entries.
- */
-int nfs4_acl_bytes(int entries)
-{
- return sizeof(struct nfs4_acl) + entries * sizeof(struct nfs4_ace);
+ return ace->e_flags & RICHACE_IDENTIFIER_GROUP ? ACL_GROUP : ACL_USER;
}

-static struct {
- char *string;
- int stringlen;
- int type;
-} s2t_map[] = {
- {
- .string = "OWNER@",
- .stringlen = sizeof("OWNER@") - 1,
- .type = NFS4_ACL_WHO_OWNER,
- },
- {
- .string = "GROUP@",
- .stringlen = sizeof("GROUP@") - 1,
- .type = NFS4_ACL_WHO_GROUP,
- },
- {
- .string = "EVERYONE@",
- .stringlen = sizeof("EVERYONE@") - 1,
- .type = NFS4_ACL_WHO_EVERYONE,
- },
-};
-
-int
-nfs4_acl_get_whotype(char *p, u32 len)
+__be32 nfsd4_decode_ace_who(struct richace *ace, struct svc_rqst *rqstp,
+ char *who, u32 len)
{
- int i;
-
- for (i = 0; i < ARRAY_SIZE(s2t_map); i++) {
- if (s2t_map[i].stringlen == len &&
- 0 == memcmp(s2t_map[i].string, p, len))
- return s2t_map[i].type;
+ int special_id;
+
+ special_id = nfs4acl_who_to_special_id(who, len);
+ if (special_id >= 0) {
+ ace->e_flags |= RICHACE_SPECIAL_WHO;
+ ace->e_flags &= ~RICHACE_IDENTIFIER_GROUP;
+ ace->e_id.special = special_id;
+ return nfs_ok;
}
- return NFS4_ACL_WHO_NAMED;
+ if (ace->e_flags & RICHACE_IDENTIFIER_GROUP)
+ return nfsd_map_name_to_gid(rqstp, who, len, &ace->e_id.gid);
+ else
+ return nfsd_map_name_to_uid(rqstp, who, len, &ace->e_id.uid);
}

-__be32 nfs4_acl_write_who(struct xdr_stream *xdr, int who)
+__be32 nfsd4_encode_ace_who(struct xdr_stream *xdr, struct svc_rqst *rqstp,
+ struct richace *ace)
{
- __be32 *p;
- int i;
-
- for (i = 0; i < ARRAY_SIZE(s2t_map); i++) {
- if (s2t_map[i].type != who)
- continue;
- p = xdr_reserve_space(xdr, s2t_map[i].stringlen + 4);
+ if (ace->e_flags & RICHACE_SPECIAL_WHO) {
+ unsigned int special_id = ace->e_id.special;
+ const char *who;
+ unsigned int len;
+ __be32 *p;
+
+ if (!nfs4acl_special_id_to_who(special_id, &who, &len)) {
+ WARN_ON_ONCE(1);
+ return nfserr_serverfault;
+ }
+ p = xdr_reserve_space(xdr, len + 4);
if (!p)
return nfserr_resource;
- p = xdr_encode_opaque(p, s2t_map[i].string,
- s2t_map[i].stringlen);
+ p = xdr_encode_opaque(p, who, len);
return 0;
}
- WARN_ON_ONCE(1);
- return nfserr_serverfault;
+ if (ace->e_flags & RICHACE_IDENTIFIER_GROUP)
+ return nfsd4_encode_group(xdr, rqstp, ace->e_id.gid);
+ else
+ return nfsd4_encode_user(xdr, rqstp, ace->e_id.uid);
}
diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 4ce6b97..2430235 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -159,12 +159,12 @@ is_create_with_attrs(struct nfsd4_open *open)
* in the returned attr bitmap.
*/
static void
-do_set_nfs4_acl(struct svc_rqst *rqstp, struct svc_fh *fhp,
- struct nfs4_acl *acl, u32 *bmval)
+do_set_acl(struct svc_rqst *rqstp, struct svc_fh *fhp, struct richacl *acl,
+ u32 *bmval)
{
__be32 status;

- status = nfsd4_set_nfs4_acl(rqstp, fhp, acl);
+ status = nfsd4_set_acl(rqstp, fhp, acl);
if (status)
/*
* We should probably fail the whole open at this point,
@@ -299,7 +299,7 @@ do_open_lookup(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stru
goto out;

if (is_create_with_attrs(open) && open->op_acl != NULL)
- do_set_nfs4_acl(rqstp, *resfh, open->op_acl, open->op_bmval);
+ do_set_acl(rqstp, *resfh, open->op_acl, open->op_bmval);

nfsd4_set_open_owner_reply_cache(cstate, open, *resfh);
accmode = NFSD_MAY_NOP;
@@ -672,8 +672,7 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
nfsd4_security_inode_setsecctx(&resfh, &create->cr_label, create->cr_bmval);

if (create->cr_acl != NULL)
- do_set_nfs4_acl(rqstp, &resfh, create->cr_acl,
- create->cr_bmval);
+ do_set_acl(rqstp, &resfh, create->cr_acl, create->cr_bmval);

fh_unlock(&cstate->current_fh);
set_change_info(&create->cr_cinfo, &cstate->current_fh);
@@ -938,8 +937,8 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
goto out;

if (setattr->sa_acl != NULL)
- status = nfsd4_set_nfs4_acl(rqstp, &cstate->current_fh,
- setattr->sa_acl);
+ status = nfsd4_set_acl(rqstp, &cstate->current_fh,
+ setattr->sa_acl);
if (status)
goto out;
if (setattr->sa_label.len)
diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index b8db5a7..8603f40 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -303,7 +303,7 @@ nfsd4_decode_bitmap(struct nfsd4_compoundargs *argp, u32 *bmval)

static __be32
nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,
- struct iattr *iattr, struct nfs4_acl **acl,
+ struct iattr *iattr, struct richacl **acl,
struct xdr_netobj *label)
{
int expected_len, len = 0;
@@ -326,38 +326,31 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,
}
if (bmval[0] & FATTR4_WORD0_ACL) {
u32 nace;
- struct nfs4_ace *ace;
+ struct richace *ace;

READ_BUF(4); len += 4;
nace = be32_to_cpup(p++);

- if (nace > NFS4_ACL_MAX)
+ if (nace > NFSD4_ACL_MAX)
return nfserr_fbig;

- *acl = svcxdr_tmpalloc(argp, nfs4_acl_bytes(nace));
+ *acl = svcxdr_alloc_richacl(argp, nace);
if (*acl == NULL)
return nfserr_jukebox;

- (*acl)->naces = nace;
- for (ace = (*acl)->aces; ace < (*acl)->aces + nace; ace++) {
+ richacl_for_each_entry(ace, *acl) {
READ_BUF(16); len += 16;
- ace->type = be32_to_cpup(p++);
- ace->flag = be32_to_cpup(p++);
- ace->access_mask = be32_to_cpup(p++);
+ ace->e_type = be32_to_cpup(p++);
+ ace->e_flags = be32_to_cpup(p++);
+ ace->e_mask = be32_to_cpup(p++);
+ if (ace->e_flags & RICHACE_SPECIAL_WHO)
+ return nfserr_inval;
dummy32 = be32_to_cpup(p++);
READ_BUF(dummy32);
len += XDR_QUADLEN(dummy32) << 2;
READMEM(buf, dummy32);
- ace->whotype = nfs4_acl_get_whotype(buf, dummy32);
- status = nfs_ok;
- if (ace->whotype != NFS4_ACL_WHO_NAMED)
- ;
- else if (ace->flag & NFS4_ACE_IDENTIFIER_GROUP)
- status = nfsd_map_name_to_gid(argp->rqstp,
- buf, dummy32, &ace->who_gid);
- else
- status = nfsd_map_name_to_uid(argp->rqstp,
- buf, dummy32, &ace->who_uid);
+ status = nfsd4_decode_ace_who(ace, argp->rqstp,
+ buf, dummy32);
if (status)
return status;
}
@@ -2148,18 +2141,6 @@ static u32 nfs4_file_type(umode_t mode)
}

static inline __be32
-nfsd4_encode_aclname(struct xdr_stream *xdr, struct svc_rqst *rqstp,
- struct nfs4_ace *ace)
-{
- if (ace->whotype != NFS4_ACL_WHO_NAMED)
- return nfs4_acl_write_who(xdr, ace->whotype);
- else if (ace->flag & NFS4_ACE_IDENTIFIER_GROUP)
- return nfsd4_encode_group(xdr, rqstp, ace->who_gid);
- else
- return nfsd4_encode_user(xdr, rqstp, ace->who_uid);
-}
-
-static inline __be32
nfsd4_encode_layout_type(struct xdr_stream *xdr, enum pnfs_layouttype layout_type)
{
__be32 *p;
@@ -2303,7 +2284,7 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
u32 rdattr_err = 0;
__be32 status;
int err;
- struct nfs4_acl *acl = NULL;
+ struct richacl *acl = NULL;
void *context = NULL;
int contextlen;
bool contextsupport = false;
@@ -2349,7 +2330,7 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
fhp = tempfh;
}
if (bmval0 & FATTR4_WORD0_ACL) {
- err = nfsd4_get_nfs4_acl(rqstp, dentry, &acl);
+ err = nfsd4_get_acl(rqstp, dentry, &acl);
if (err == -EOPNOTSUPP)
bmval0 &= ~FATTR4_WORD0_ACL;
else if (err == -EINVAL) {
@@ -2504,7 +2485,7 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
*p++ = cpu_to_be32(rdattr_err);
}
if (bmval0 & FATTR4_WORD0_ACL) {
- struct nfs4_ace *ace;
+ struct richace *ace;

if (acl == NULL) {
p = xdr_reserve_space(xdr, 4);
@@ -2517,17 +2498,16 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp,
p = xdr_reserve_space(xdr, 4);
if (!p)
goto out_resource;
- *p++ = cpu_to_be32(acl->naces);
+ *p++ = cpu_to_be32(acl->a_count);

- for (ace = acl->aces; ace < acl->aces + acl->naces; ace++) {
+ richacl_for_each_entry(ace, acl) {
p = xdr_reserve_space(xdr, 4*3);
if (!p)
goto out_resource;
- *p++ = cpu_to_be32(ace->type);
- *p++ = cpu_to_be32(ace->flag);
- *p++ = cpu_to_be32(ace->access_mask &
- NFS4_ACE_MASK_ALL);
- status = nfsd4_encode_aclname(xdr, rqstp, ace);
+ *p++ = cpu_to_be32(ace->e_type);
+ *p++ = cpu_to_be32(ace->e_flags & ~RICHACE_SPECIAL_WHO);
+ *p++ = cpu_to_be32(ace->e_mask & NFS4_ACE_MASK_ALL);
+ status = nfsd4_encode_ace_who(xdr, rqstp, ace);
if (status)
goto out;
}
@@ -2792,7 +2772,7 @@ out:
if (context)
security_release_secctx(context, contextlen);
#endif /* CONFIG_NFSD_V4_SECURITY_LABEL */
- kfree(acl);
+ richacl_put(acl);
if (tempfh) {
fh_put(tempfh);
kfree(tempfh);
diff --git a/fs/nfsd/xdr4.h b/fs/nfsd/xdr4.h
index b698585..c311066 100644
--- a/fs/nfsd/xdr4.h
+++ b/fs/nfsd/xdr4.h
@@ -118,7 +118,7 @@ struct nfsd4_create {
u32 cr_bmval[3]; /* request */
struct iattr cr_iattr; /* request */
struct nfsd4_change_info cr_cinfo; /* response */
- struct nfs4_acl *cr_acl;
+ struct richacl *cr_acl;
struct xdr_netobj cr_label;
};
#define cr_datalen u.link.datalen
@@ -248,7 +248,7 @@ struct nfsd4_open {
struct nfs4_file *op_file; /* used during processing */
struct nfs4_ol_stateid *op_stp; /* used during processing */
struct nfs4_clnt_odstate *op_odstate; /* used during processing */
- struct nfs4_acl *op_acl;
+ struct richacl *op_acl;
struct xdr_netobj op_label;
};

@@ -332,7 +332,7 @@ struct nfsd4_setattr {
stateid_t sa_stateid; /* request */
u32 sa_bmval[3]; /* request */
struct iattr sa_iattr; /* request */
- struct nfs4_acl *sa_acl;
+ struct richacl *sa_acl;
struct xdr_netobj sa_label;
};

diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h
index 00121f2..1422fc6 100644
--- a/include/linux/nfs4.h
+++ b/include/linux/nfs4.h
@@ -16,29 +16,6 @@
#include <linux/uidgid.h>
#include <uapi/linux/nfs4.h>

-enum nfs4_acl_whotype {
- NFS4_ACL_WHO_NAMED = 0,
- NFS4_ACL_WHO_OWNER,
- NFS4_ACL_WHO_GROUP,
- NFS4_ACL_WHO_EVERYONE,
-};
-
-struct nfs4_ace {
- uint32_t type;
- uint32_t flag;
- uint32_t access_mask;
- int whotype;
- union {
- kuid_t who_uid;
- kgid_t who_gid;
- };
-};
-
-struct nfs4_acl {
- uint32_t naces;
- struct nfs4_ace aces[0];
-};
-
#define NFS4_MAXLABELLEN 2048

struct nfs4_label {
diff --git a/include/linux/nfs4acl.h b/include/linux/nfs4acl.h
new file mode 100644
index 0000000..db9f9a6
--- /dev/null
+++ b/include/linux/nfs4acl.h
@@ -0,0 +1,7 @@
+#ifndef __LINUX_NFS4ACL_H
+#define __LINUX_NFS4ACL_H
+
+int nfs4acl_who_to_special_id(const char *, u32);
+bool nfs4acl_special_id_to_who(unsigned int, const char **, unsigned int *);
+
+#endif
--
2.4.3


2015-09-27 22:12:26

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 28/41] nfsd: Keep list of acls to dispose of in compoundargs

We will decode acls in requests into richacls. Even if unlikely, there
can be more than one acl in a single request; those richacls need to be
richacl_put() at the end of the request instead of kfree()d, so keep a
list of acls in compoundargs for that.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Acked-by: J. Bruce Fields <[email protected]>
---
fs/nfsd/nfs4xdr.c | 27 +++++++++++++++++++++++++++
fs/nfsd/xdr4.h | 6 ++++++
2 files changed, 33 insertions(+)

diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
index 51c9e9c..b8db5a7 100644
--- a/fs/nfsd/nfs4xdr.c
+++ b/fs/nfsd/nfs4xdr.c
@@ -40,6 +40,7 @@
#include <linux/utsname.h>
#include <linux/pagemap.h>
#include <linux/sunrpc/svcauth_gss.h>
+#include <linux/richacl.h>

#include "idmap.h"
#include "acl.h"
@@ -196,6 +197,24 @@ svcxdr_tmpalloc(struct nfsd4_compoundargs *argp, u32 len)
return tb->buf;
}

+static struct richacl *
+svcxdr_alloc_richacl(struct nfsd4_compoundargs *argp, u32 nace)
+{
+ struct svcxdr_richacl *acls;
+
+ acls = kmalloc(sizeof(*acls), GFP_KERNEL);
+ if (!acls)
+ return NULL;
+ acls->acl = richacl_alloc(nace, GFP_KERNEL);
+ if (!acls->acl) {
+ kfree(acls);
+ return NULL;
+ }
+ acls->next = argp->acls;
+ argp->acls = acls;
+ return acls->acl;
+}
+
/*
* For xdr strings that need to be passed to other kernel api's
* as null-terminated strings.
@@ -4437,6 +4456,13 @@ int nfsd4_release_compoundargs(void *rq, __be32 *p, void *resp)
args->to_free = tb->next;
kfree(tb);
}
+ while (args->acls) {
+ struct svcxdr_richacl *acls = args->acls;
+
+ args->acls = acls->next;
+ richacl_put(acls->acl);
+ kfree(acls);
+ }
return 1;
}

@@ -4455,6 +4481,7 @@ nfs4svc_decode_compoundargs(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_comp
args->pagelen = rqstp->rq_arg.page_len;
args->tmpp = NULL;
args->to_free = NULL;
+ args->acls = NULL;
args->ops = args->iops;
args->rqstp = rqstp;

diff --git a/fs/nfsd/xdr4.h b/fs/nfsd/xdr4.h
index 9f99100..b698585 100644
--- a/fs/nfsd/xdr4.h
+++ b/fs/nfsd/xdr4.h
@@ -570,6 +570,11 @@ struct svcxdr_tmpbuf {
char buf[];
};

+struct svcxdr_richacl {
+ struct svcxdr_richacl *next;
+ struct richacl *acl;
+};
+
struct nfsd4_compoundargs {
/* scratch variables for XDR decode */
__be32 * p;
@@ -579,6 +584,7 @@ struct nfsd4_compoundargs {
__be32 tmp[8];
__be32 * tmpp;
struct svcxdr_tmpbuf *to_free;
+ struct svcxdr_richacl *acls;

struct svc_rqst *rqstp;

--
2.4.3


2015-09-27 22:12:21

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 27/41] richacl: Create richacl from mode values

A file can have "no acl" in the sense that only the file mode permission
bits determine access. In that case, the getxattr system call fails with
errno == ENODATA (No such attribute).

Over the NFSv4 protocol, a file always has an acl, and we convert the file
mode permission bits into an equivalent acl with richacl_from_mode. Such
"trivial" acls can be converted back to a file mode with
richacl_equiv_mode.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_compat.c | 88 +++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 1 +
2 files changed, 89 insertions(+)

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index c99274f..df807c0 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -827,3 +827,91 @@ richacl_apply_masks(struct richacl **acl, kuid_t owner)
return 0;
}
EXPORT_SYMBOL_GPL(richacl_apply_masks);
+
+/**
+ * richacl_from_mode - create an acl which corresponds to @mode
+ *
+ * The resulting acl doesn't have the RICHACL_MASKED flag set.
+ *
+ * @mode: file mode including the file type
+ */
+struct richacl *
+richacl_from_mode(mode_t mode)
+{
+ unsigned int owner_mask = richacl_mode_to_mask(mode >> 6);
+ unsigned int group_mask = richacl_mode_to_mask(mode >> 3);
+ unsigned int other_mask = richacl_mode_to_mask(mode);
+ unsigned int denied;
+ unsigned int entries = 0;
+ struct richacl *acl;
+ struct richace *ace;
+
+ /* RICHACE_DELETE_CHILD is meaningless for non-directories. */
+ if (!S_ISDIR(mode)) {
+ owner_mask &= ~RICHACE_DELETE_CHILD;
+ group_mask &= ~RICHACE_DELETE_CHILD;
+ other_mask &= ~RICHACE_DELETE_CHILD;
+ }
+
+ denied = ~owner_mask & (group_mask | other_mask);
+ if (denied)
+ entries++; /* owner@ deny entry needed */
+ if (owner_mask & ~(group_mask & other_mask))
+ entries++; /* owner@ allow entry needed */
+ denied = ~group_mask & other_mask;
+ if (denied)
+ entries++; /* group@ deny entry needed */
+ if (group_mask & ~other_mask)
+ entries++; /* group@ allow entry needed */
+ if (other_mask)
+ entries++; /* everyone@ allow entry needed */
+
+ acl = richacl_alloc(entries, GFP_KERNEL);
+ if (!acl)
+ return NULL;
+ acl->a_owner_mask = owner_mask;
+ acl->a_group_mask = group_mask;
+ acl->a_other_mask = other_mask;
+ ace = acl->a_entries;
+
+ denied = ~owner_mask & (group_mask | other_mask);
+ if (denied) {
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = denied;
+ ace->e_id.special = RICHACE_OWNER_SPECIAL_ID;
+ ace++;
+ }
+ if (owner_mask & ~(group_mask & other_mask)) {
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = owner_mask;
+ ace->e_id.special = RICHACE_OWNER_SPECIAL_ID;
+ ace++;
+ }
+ denied = ~group_mask & other_mask;
+ if (denied) {
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = denied;
+ ace->e_id.special = RICHACE_GROUP_SPECIAL_ID;
+ ace++;
+ }
+ if (group_mask & ~other_mask) {
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = group_mask;
+ ace->e_id.special = RICHACE_GROUP_SPECIAL_ID;
+ ace++;
+ }
+ if (other_mask) {
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = other_mask;
+ ace->e_id.special = RICHACE_EVERYONE_SPECIAL_ID;
+ ace++;
+ }
+
+ return acl;
+}
+EXPORT_SYMBOL_GPL(richacl_from_mode);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index a945f3c..f48c04e 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -334,5 +334,6 @@ extern struct richacl *richacl_create(struct inode *, struct inode *);

/* richacl_compat.c */
extern int richacl_apply_masks(struct richacl **, kuid_t);
+extern struct richacl *richacl_from_mode(mode_t);

#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:12:15

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 26/41] richacl: Apply the file masks to a richacl

Put all the pieces of the acl transformation puzzle together for
computing a richacl which has the file masks "applied" so that the
standard nfsv4 access check algorithm can be used on the richacl.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_compat.c | 101 ++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 3 ++
2 files changed, 104 insertions(+)

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index 7e25343..c99274f 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -726,3 +726,104 @@ richacl_isolate_group_class(struct richacl_alloc *alloc, unsigned int deny)
}
return 0;
}
+
+/**
+ * __richacl_apply_masks - apply the file masks to all aces
+ * @alloc: acl and number of allocated entries
+ *
+ * Apply the owner mask to owner@ aces, the other mask to
+ * everyone@ aces, and the group mask to all other aces.
+ *
+ * The previous transformations have brought the acl into a
+ * form in which applying the masks will not lead to the
+ * accidental loss of permissions anymore.
+ */
+static int
+__richacl_apply_masks(struct richacl_alloc *alloc, kuid_t owner)
+{
+ struct richace *ace;
+
+ richacl_for_each_entry(ace, alloc->acl) {
+ unsigned int mask;
+
+ if (richace_is_inherit_only(ace) || !richace_is_allow(ace))
+ continue;
+ if (richace_is_owner(ace) ||
+ (richace_is_unix_user(ace) && uid_eq(owner, ace->e_id.uid)))
+ mask = alloc->acl->a_owner_mask;
+ else if (richace_is_everyone(ace))
+ mask = alloc->acl->a_other_mask;
+ else
+ mask = alloc->acl->a_group_mask;
+ if (richace_change_mask(alloc, &ace, ace->e_mask & mask))
+ return -1;
+ }
+ return 0;
+}
+
+/**
+ * richacl_apply_masks - apply the masks to the acl
+ *
+ * Transform @acl so that the standard NFSv4 permission check algorithm (which
+ * is not aware of file masks) will compute the same access decisions as the
+ * richacl permission check algorithm (which looks at the acl and the file
+ * masks).
+ *
+ * This algorithm is split into several steps:
+ *
+ * - Move everyone@ aces to the end of the acl. This simplifies the other
+ * transformations, and allows the everyone@ allow ace at the end of the
+ * acl to eventually allow permissions to the other class only.
+ *
+ * - Propagate everyone@ permissions up the acl. This transformation makes
+ * sure that the owner and group class aces won't lose any permissions when
+ * we apply the other mask to the everyone@ allow ace at the end of the acl.
+ *
+ * - Apply the file masks to all aces.
+ *
+ * - Make sure everyone is granted the other mask permissions. This step can
+ * elevate elevate permissions for the owner and group classes, which is
+ * corrected later.
+ *
+ * - Make sure that the group class is not granted any permissions from
+ * everyone@.
+ *
+ * - Make sure the owner is granted the owner mask permissions.
+ *
+ * - Make sure the owner is not granted any permissions beyond the owner
+ * mask from group class aces or from everyone@.
+ *
+ * NOTE: Depending on the acl and file masks, this algorithm can increase the
+ * number of aces by almost a factor of three in the worst case. This may make
+ * the acl too large for some purposes.
+ */
+int
+richacl_apply_masks(struct richacl **acl, kuid_t owner)
+{
+ if ((*acl)->a_flags & RICHACL_MASKED) {
+ struct richacl_alloc alloc = {
+ .acl = richacl_clone(*acl, GFP_KERNEL),
+ .count = (*acl)->a_count,
+ };
+ unsigned int added = 0;
+
+ if (!alloc.acl)
+ return -ENOMEM;
+ if (richacl_move_everyone_aces_down(&alloc) ||
+ richacl_propagate_everyone(&alloc) ||
+ __richacl_apply_masks(&alloc, owner) ||
+ richacl_set_other_permissions(&alloc, &added) ||
+ richacl_isolate_group_class(&alloc, added) ||
+ richacl_set_owner_permissions(&alloc) ||
+ richacl_isolate_owner_class(&alloc)) {
+ richacl_put(alloc.acl);
+ return -ENOMEM;
+ }
+
+ alloc.acl->a_flags &= ~(RICHACL_WRITE_THROUGH | RICHACL_MASKED);
+ richacl_put(*acl);
+ *acl = alloc.acl;
+ }
+ return 0;
+}
+EXPORT_SYMBOL_GPL(richacl_apply_masks);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index 832b06c..a945f3c 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -332,4 +332,7 @@ extern struct richacl *richacl_inherit(const struct richacl *, int);
extern int richacl_permission(struct inode *, const struct richacl *, int);
extern struct richacl *richacl_create(struct inode *, struct inode *);

+/* richacl_compat.c */
+extern int richacl_apply_masks(struct richacl **, kuid_t);
+
#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:12:09

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 25/41] richacl: Isolate the owner and group classes

When applying the file masks to an acl, we need to ensure that no
process gets more permissions than allowed by its file mask.

This may require inserting an owner@ deny ace to ensure this if the
owner mask contains fewer permissions than the group or other mask. For
example, when applying mode 0446 to the following acl:

everyone@:rw::allow

A deny ace needs to be inserted so that the owner won't get elevated
write access:

owner@:w::deny
everyone@:rw::allow

Likewise, we may need to insert group class deny aces if the group mask
contains fewer permissions than the other mask. For example, when
applying mode 0646 to the following acl:

owner@:rw::allow
everyone@:rw::allow

A deny ace needs to be inserted so that the owning group won't get
elevated write access:

owner@:rw::allow
group@:w::deny
everyone@:rw::allow

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_compat.c | 223 ++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 223 insertions(+)

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index 21ee31e..7e25343 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -503,3 +503,226 @@ richacl_set_other_permissions(struct richacl_alloc *alloc, unsigned int *added)
}
return 0;
}
+
+/**
+ * richacl_max_allowed - maximum permissions that anybody is allowed
+ */
+static unsigned int
+richacl_max_allowed(struct richacl *acl)
+{
+ struct richace *ace;
+ unsigned int allowed = 0;
+
+ richacl_for_each_entry_reverse(ace, acl) {
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_allow(ace))
+ allowed |= ace->e_mask;
+ else if (richace_is_deny(ace)) {
+ if (richace_is_everyone(ace))
+ allowed &= ~ace->e_mask;
+ }
+ }
+ return allowed;
+}
+
+/**
+ * richacl_isolate_owner_class - limit the owner class to the owner file mask
+ * @alloc: acl and number of allocated entries
+ *
+ * POSIX requires that after a chmod, the owner class is granted no more
+ * permissions than the owner file permission bits. For richacls, this
+ * means that the owner class must not be granted any permissions that the
+ * owner mask does not include.
+ *
+ * When we apply file masks to an acl which grant more permissions to the group
+ * or other class than to the owner class, we may end up in a situation where
+ * the owner is granted additional permissions from other aces. For example,
+ * given this acl:
+ *
+ * everyone@:rwx::allow
+ *
+ * when file masks corresponding to mode 0406 are applied, after
+ * richacl_propagate_everyone() and __richacl_apply_masks(), we end up with:
+ *
+ * owner@:r::allow
+ * everyone@:rw::allow
+ *
+ * This acl still grants the owner rw access through the everyone@ allow ace.
+ * To fix this, we must deny the owner w access:
+ *
+ * owner@:w::deny
+ * owner@:r::allow
+ * everyone@:rw::allow
+ */
+static int
+richacl_isolate_owner_class(struct richacl_alloc *alloc)
+{
+ struct richacl *acl = alloc->acl;
+ struct richace *ace;
+ unsigned int deny;
+
+ deny = richacl_max_allowed(acl) & ~acl->a_owner_mask;
+ if (!deny)
+ return 0;
+
+ /*
+ * Figure out if we can update an existig OWNER@ DENY entry.
+ */
+ richacl_for_each_entry(ace, acl) {
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_allow(ace))
+ break;
+ if (richace_is_owner(ace))
+ return richace_change_mask(alloc, &ace,
+ ace->e_mask | deny);
+ }
+
+ /* Insert an owner@ deny entry at the front. */
+ ace = acl->a_entries;
+ if (richacl_insert_entry(alloc, &ace))
+ return -1;
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = deny;
+ ace->e_id.special = RICHACE_OWNER_SPECIAL_ID;
+ return 0;
+}
+
+/**
+ * __richacl_isolate_who - isolate entry from everyone@ allow entry
+ * @alloc: acl and number of allocated entries
+ * @who: identifier to isolate
+ * @deny: permissions this identifier should not be allowed
+ *
+ * See richacl_isolate_group_class().
+ */
+static int
+__richacl_isolate_who(struct richacl_alloc *alloc, struct richace *who,
+ unsigned int deny)
+{
+ struct richacl *acl = alloc->acl;
+ struct richace *ace, who_copy;
+ int n;
+
+ /*
+ * Compute the permissions already defined for @who. There are no
+ * everyone@ deny aces left in the acl at this stage.
+ */
+ richacl_for_each_entry(ace, acl) {
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_same_identifier(ace, who))
+ deny &= ~ace->e_mask;
+ }
+ if (!deny)
+ return 0;
+
+ /*
+ * Figure out if we can update an existig deny entry. Start from the
+ * entry before the trailing everyone@ allow entry. We will not hit
+ * everyone@ entries in the loop.
+ */
+ for (n = acl->a_count - 2; n != -1; n--) {
+ ace = acl->a_entries + n;
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_deny(ace)) {
+ if (richace_is_same_identifier(ace, who))
+ return richace_change_mask(alloc, &ace,
+ ace->e_mask | deny);
+ } else if (richace_is_allow(ace) &&
+ (ace->e_mask & deny))
+ break;
+ }
+
+ /*
+ * Insert a new entry before the trailing everyone@ deny entry.
+ */
+ richace_copy(&who_copy, who);
+ ace = acl->a_entries + acl->a_count - 1;
+ if (richacl_insert_entry(alloc, &ace))
+ return -1;
+ richace_copy(ace, &who_copy);
+ ace->e_type = RICHACE_ACCESS_DENIED_ACE_TYPE;
+ ace->e_flags &= ~RICHACE_INHERITANCE_FLAGS;
+ ace->e_mask = deny;
+ return 0;
+}
+
+/**
+ * richacl_isolate_group_class - limit the group class to the group file mask
+ * @alloc: acl and number of allocated entries
+ * @deny: additional permissions to deny
+ *
+ * POSIX requires that after a chmod, the group class is granted no more
+ * permissions than the group file permission bits. For richacls, this
+ * means that the group class must not be granted any permissions that the
+ * group mask does not include.
+ *
+ * When we apply file masks to an acl which grant more permissions to the other
+ * class than to the group class, we may end up in a situation where processes
+ * in the group class are granted additional permission from other aces. For
+ * example, given this acl:
+ *
+ * joe:rwx::allow
+ * everyone@:rwx::allow
+ *
+ * when file masks corresponding to mode 0646 are applied, after
+ * richacl_propagate_everyone() and __richacl_apply_masks(), we end up with:
+ *
+ * joe:r::allow
+ * owner@:rw::allow
+ * group@:r::allow
+ * everyone@:rw::allow
+ *
+ * This acl still grants joe and group@ rw access through the everyone@ allow
+ * ace. To fix this, we must deny w access to group class aces before the
+ * everyone@ allow ace at the end of the acl:
+ *
+ * joe:r::allow
+ * owner@:rw::allow
+ * group@:r::allow
+ * joe:w::deny
+ * group@:w::deny
+ * everyone@:rw::allow
+ */
+static int
+richacl_isolate_group_class(struct richacl_alloc *alloc, unsigned int deny)
+{
+ struct richace who = {
+ .e_flags = RICHACE_SPECIAL_WHO,
+ .e_id.special = RICHACE_GROUP_SPECIAL_ID,
+ };
+ struct richace *ace;
+
+ if (!alloc->acl->a_count)
+ return 0;
+ ace = alloc->acl->a_entries + alloc->acl->a_count - 1;
+ if (richace_is_inherit_only(ace) || !richace_is_everyone(ace))
+ return 0;
+ deny |= ace->e_mask & ~alloc->acl->a_group_mask;
+
+ if (deny) {
+ unsigned int n;
+
+ if (__richacl_isolate_who(alloc, &who, deny))
+ return -1;
+ /*
+ * Start from the entry before the trailing everyone@ allow
+ * entry. We will not hit everyone@ entries in the loop.
+ */
+ for (n = alloc->acl->a_count - 2; n != -1; n--) {
+ ace = alloc->acl->a_entries + n;
+
+ if (richace_is_inherit_only(ace) ||
+ richace_is_owner(ace) ||
+ richace_is_group(ace))
+ continue;
+ if (__richacl_isolate_who(alloc, ace, deny))
+ return -1;
+ }
+ }
+ return 0;
+}
--
2.4.3


2015-09-27 22:12:03

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 24/41] richacl: Set the other permissions to the other mask

Change the acl so that everyone@ is granted the permissions set in the
other mask.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_compat.c | 41 +++++++++++++++++++++++++++++++++++++++++
1 file changed, 41 insertions(+)

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index 16deeb4..21ee31e 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -462,3 +462,44 @@ richacl_set_owner_permissions(struct richacl_alloc *alloc)
}
return 0;
}
+
+/**
+ * richacl_set_other_permissions - set the other permissions to the other mask
+ * @alloc: acl and number of allocated entries
+ * @added: permissions added for everyone@
+ *
+ * Change the acl so that everyone@ is granted the permissions set in the other
+ * mask. This leaves at most one efective everyone@ allow entry at the end of
+ * the acl. If everyone@ end up being granted additional permissions, these
+ * permissions are returned in @added.
+ */
+static int
+richacl_set_other_permissions(struct richacl_alloc *alloc, unsigned int *added)
+{
+ struct richacl *acl = alloc->acl;
+ unsigned int x = RICHACE_POSIX_ALWAYS_ALLOWED;
+ unsigned int other_mask = acl->a_other_mask & ~x;
+ struct richace *ace;
+
+ if (!(other_mask &&
+ (acl->a_flags & RICHACL_WRITE_THROUGH)))
+ return 0;
+
+ *added = other_mask;
+ ace = acl->a_entries + acl->a_count - 1;
+ if (acl->a_count == 0 ||
+ !richace_is_everyone(ace) ||
+ richace_is_inherit_only(ace)) {
+ ace++;
+ if (richacl_insert_entry(alloc, &ace))
+ return -1;
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = other_mask;
+ ace->e_id.special = RICHACE_EVERYONE_SPECIAL_ID;
+ } else {
+ *added &= ~ace->e_mask;
+ richace_change_mask(alloc, &ace, other_mask);
+ }
+ return 0;
+}
--
2.4.3


2015-09-27 22:11:58

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 23/41] richacl: Set the owner permissions to the owner mask

In the write-through case, change the acl so that owner@ is granted the
permissions set in the owner mask (to match what the permission check
algorithm grants the owner).

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_compat.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 46 insertions(+)

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index 7d007d7..16deeb4 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -416,3 +416,49 @@ richacl_propagate_everyone(struct richacl_alloc *alloc)
}
return 0;
}
+
+/**
+ * richacl_set_owner_permissions - set the owner permissions to the owner mask
+ *
+ * In the write-through case, change the acl so that owner@ is granted the
+ * permissions set in the owner mask (to match what the permission check
+ * algorithm grants the owner). This leaves at most one efective owner@ allow
+ * entry at the beginning of the acl.
+ */
+static int
+richacl_set_owner_permissions(struct richacl_alloc *alloc)
+{
+ unsigned int x = RICHACE_POSIX_ALWAYS_ALLOWED;
+ unsigned int owner_mask = alloc->acl->a_owner_mask & ~x;
+ unsigned int denied = 0;
+ struct richace *ace;
+
+ if (!((alloc->acl->a_flags & RICHACL_WRITE_THROUGH)))
+ return 0;
+
+ richacl_for_each_entry(ace, alloc->acl) {
+ if (richace_is_owner(ace)) {
+ if (richace_is_allow(ace) && !(owner_mask & denied)) {
+ richace_change_mask(alloc, &ace, owner_mask);
+ owner_mask = 0;
+ } else
+ richace_change_mask(alloc, &ace, 0);
+ } else {
+ if (richace_is_deny(ace))
+ denied |= ace->e_mask;
+ }
+ }
+
+ if (owner_mask & (denied |
+ ~alloc->acl->a_other_mask |
+ ~alloc->acl->a_group_mask)) {
+ ace = alloc->acl->a_entries;
+ if (richacl_insert_entry(alloc, &ace))
+ return -1;
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = owner_mask;
+ ace->e_id.special = RICHACE_OWNER_SPECIAL_ID;
+ }
+ return 0;
+}
--
2.4.3


2015-09-27 22:11:52

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 22/41] richacl: Propagate everyone@ permissions to other aces

The trailing everyone@ allow ace can grant permissions to all file
classes including the owner and group class. Before we can apply the
other mask to this entry to turn it into an "other class" entry, we need
to ensure that members of the owner or group class will not lose any
permissions from that ace.

Conceptually, we do this by inserting additional <who>:<allow>::allow
entries before the trailing everyone@ allow ace with the same
permissions as the trailing everyone@ allow ace for owner@, group@, and
all explicitly mentioned users and groups. (In practice, we will rarely
need to insert any additional aces in this step.)

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_compat.c | 198 ++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 198 insertions(+)

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index 4f0acf5..7d007d7 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -218,3 +218,201 @@ richacl_move_everyone_aces_down(struct richacl_alloc *alloc)
}
return 0;
}
+
+/**
+ * __richacl_propagate_everyone - propagate everyone@ permissions up for @who
+ * @alloc: acl and number of allocated entries
+ * @who: identifier to propagate permissions for
+ * @allow: permissions to propagate up
+ *
+ * Propagate the permissions in @allow up from the end of the acl to the start
+ * for the specified principal @who.
+ *
+ * The simplest possible approach to achieve this would be to insert a
+ * "<who>:<allow>::allow" ace before the final everyone@ allow ace. Since this
+ * would often result in aces which are not needed or which could be merged
+ * with an existing ace, we make the following optimizations:
+ *
+ * - We go through the acl and determine which permissions are already
+ * allowed or denied to @who, and we remove those permissions from
+ * @allow.
+ *
+ * - If the acl contains an allow ace for @who and no aces after this entry
+ * deny permissions in @allow, we add the permissions in @allow to this
+ * ace. (Propagating permissions across a deny ace which can match the
+ * process can elevate permissions.)
+ *
+ * This transformation does not alter the permissions that the acl grants.
+ */
+static int
+__richacl_propagate_everyone(struct richacl_alloc *alloc, struct richace *who,
+ unsigned int allow)
+{
+ struct richace *allow_last = NULL, *ace;
+ struct richacl *acl = alloc->acl;
+
+ /*
+ * Remove the permissions from allow that are already determined for
+ * this who value, and figure out if there is an allow entry for
+ * this who value that is "reachable" from the trailing everyone@
+ * allow ace.
+ */
+ richacl_for_each_entry(ace, acl) {
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_allow(ace)) {
+ if (richace_is_same_identifier(ace, who)) {
+ allow &= ~ace->e_mask;
+ allow_last = ace;
+ }
+ } else if (richace_is_deny(ace)) {
+ if (richace_is_same_identifier(ace, who))
+ allow &= ~ace->e_mask;
+ else if (allow & ace->e_mask)
+ allow_last = NULL;
+ }
+ }
+ ace--;
+
+ /*
+ * If for group class entries, all the remaining permissions will
+ * remain granted by the trailing everyone@ allow ace, no additional
+ * entry is needed.
+ */
+ if (!richace_is_owner(who) &&
+ richace_is_everyone(ace) &&
+ !(allow & ~(ace->e_mask & acl->a_other_mask)))
+ allow = 0;
+
+ if (allow) {
+ if (allow_last)
+ return richace_change_mask(alloc, &allow_last,
+ allow_last->e_mask | allow);
+ else {
+ struct richace who_copy;
+
+ richace_copy(&who_copy, who);
+ if (richacl_insert_entry(alloc, &ace))
+ return -1;
+ richace_copy(ace, &who_copy);
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags &= ~RICHACE_INHERITANCE_FLAGS;
+ ace->e_mask = allow;
+ }
+ }
+ return 0;
+}
+
+/**
+ * richacl_propagate_everyone - propagate everyone@ permissions up the acl
+ * @alloc: acl and number of allocated entries
+ *
+ * Make sure that group@ and all other users and groups mentioned in the acl
+ * will not lose any permissions when finally applying the other mask to the
+ * everyone@ allow ace at the end of the acl. We modify the permissions of
+ * existing entries or add new entries before the final everyone@ allow ace to
+ * achieve that.
+ *
+ * For example, the following acl implicitly grants everyone rwpx access:
+ *
+ * joe:r::allow
+ * everyone@:rwpx::allow
+ *
+ * When applying mode 0660 to this acl, group@ would lose rwp access, and joe
+ * would lose wp access even though the mode does not exclude those
+ * permissions. After propagating the everyone@ permissions, the result for
+ * applying mode 0660 becomes:
+ *
+ * owner@:rwp::allow
+ * joe:rwp::allow
+ * group@:rwp::allow
+ *
+ * Deny aces complicate the matter. For example, the following acl grants
+ * everyone but joe write access:
+ *
+ * joe:wp::deny
+ * everyone@:rwpx::allow
+ *
+ * When applying mode 0660 to this acl, group@ would lose rwp access, and joe
+ * would lose r access. After propagating the everyone@ permissions, the
+ * result for applying mode 0660 becomes:
+ *
+ * owner@:rwp::allow
+ * joe:w::deny
+ * group@:rwp::allow
+ * joe:r::allow
+ */
+static int
+richacl_propagate_everyone(struct richacl_alloc *alloc)
+{
+ struct richace who = { .e_flags = RICHACE_SPECIAL_WHO };
+ struct richacl *acl = alloc->acl;
+ struct richace *ace;
+ unsigned int owner_allow, group_allow;
+
+ if (!acl->a_count)
+ return 0;
+ ace = acl->a_entries + acl->a_count - 1;
+ if (richace_is_inherit_only(ace) || !richace_is_everyone(ace))
+ return 0;
+
+ /*
+ * Permissions the owner and group class are granted through the
+ * trailing everyone@ allow ace.
+ */
+ owner_allow = ace->e_mask & acl->a_owner_mask;
+ group_allow = ace->e_mask & acl->a_group_mask;
+
+ /*
+ * If the group or other masks hide permissions which the owner should
+ * be allowed, we need to propagate those permissions up. Otherwise,
+ * those permissions may be lost when applying the other mask to the
+ * trailing everyone@ allow ace, or when isolating the group class from
+ * the other class through additional deny aces.
+ */
+ if (owner_allow & ~(acl->a_group_mask & acl->a_other_mask)) {
+ /* Propagate everyone@ permissions through to owner@. */
+ who.e_id.special = RICHACE_OWNER_SPECIAL_ID;
+ if (__richacl_propagate_everyone(alloc, &who, owner_allow))
+ return -1;
+ acl = alloc->acl;
+ }
+
+ /*
+ * If the other mask hides permissions which the group class should be
+ * allowed, we need to propagate those permissions up to the owning
+ * group and to all other members in the group class.
+ */
+ if (group_allow & ~acl->a_other_mask) {
+ int n;
+
+ /* Propagate everyone@ permissions through to group@. */
+ who.e_id.special = RICHACE_GROUP_SPECIAL_ID;
+ if (__richacl_propagate_everyone(alloc, &who, group_allow))
+ return -1;
+ acl = alloc->acl;
+
+ /*
+ * Start from the entry before the trailing everyone@ allow
+ * entry. We will not hit everyone@ entries in the loop.
+ */
+ for (n = acl->a_count - 2; n != -1; n--) {
+ ace = acl->a_entries + n;
+
+ if (richace_is_inherit_only(ace) ||
+ richace_is_owner(ace) ||
+ richace_is_group(ace))
+ continue;
+
+ /*
+ * Any inserted entry will end up below the current
+ * entry.
+ */
+ if (__richacl_propagate_everyone(alloc, ace,
+ group_allow))
+ return -1;
+ acl = alloc->acl;
+ }
+ }
+ return 0;
+}
--
2.4.3


2015-09-27 22:11:46

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 21/41] richacl: Move everyone@ aces down the acl

The POSIX standard puts processes which are not the owner or a member in
the owning group or which match any ace other then everyone@ on the
other file class. We only know if a process is in the other class after
processing the entire acl.

Move all everyone@ aces in the acl down in the acl so that at most a
single everyone@ allow ace remains at the end. Permissions which are
not explicitly allowed are implicitly denied, so an everyone@ deny ace
is unneeded.

The everyone@ aces can be moved down the acl without changing the
permissions that the acl grants. This transformation simplifies the
following algorithms, and eventually allows us to turn the final
everyone@ allow ace into an entry for the other class.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_compat.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 65 insertions(+)

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
index 341e429..4f0acf5 100644
--- a/fs/richacl_compat.c
+++ b/fs/richacl_compat.c
@@ -153,3 +153,68 @@ richace_change_mask(struct richacl_alloc *alloc, struct richace **ace,
}
return 0;
}
+
+/**
+ * richacl_move_everyone_aces_down - move everyone@ aces to the end of the acl
+ * @alloc: acl and number of allocated entries
+ *
+ * Move all everyone aces to the end of the acl so that only a single everyone@
+ * allow ace remains at the end, and update the mask fields of all aces on the
+ * way. The last ace of the resulting acl will be an everyone@ allow ace only
+ * if @acl grants any permissions to @everyone. No @everyone deny aces will
+ * remain.
+ *
+ * This transformation does not alter the permissions that the acl grants.
+ * Having at most one everyone@ allow ace at the end of the acl helps us in the
+ * following algorithms.
+ */
+static int
+richacl_move_everyone_aces_down(struct richacl_alloc *alloc)
+{
+ struct richace *ace;
+ unsigned int allowed = 0, denied = 0;
+
+ richacl_for_each_entry(ace, alloc->acl) {
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_everyone(ace)) {
+ if (richace_is_allow(ace))
+ allowed |= (ace->e_mask & ~denied);
+ else if (richace_is_deny(ace))
+ denied |= (ace->e_mask & ~allowed);
+ else
+ continue;
+ if (richace_change_mask(alloc, &ace, 0))
+ return -1;
+ } else {
+ if (richace_is_allow(ace)) {
+ if (richace_change_mask(alloc, &ace, allowed |
+ (ace->e_mask & ~denied)))
+ return -1;
+ } else if (richace_is_deny(ace)) {
+ if (richace_change_mask(alloc, &ace, denied |
+ (ace->e_mask & ~allowed)))
+ return -1;
+ }
+ }
+ }
+ if (allowed & ~RICHACE_POSIX_ALWAYS_ALLOWED) {
+ struct richace *last_ace = ace - 1;
+
+ if (alloc->acl->a_entries &&
+ richace_is_everyone(last_ace) &&
+ richace_is_allow(last_ace) &&
+ richace_is_inherit_only(last_ace) &&
+ last_ace->e_mask == allowed)
+ last_ace->e_flags &= ~RICHACE_INHERIT_ONLY_ACE;
+ else {
+ if (richacl_insert_entry(alloc, &ace))
+ return -1;
+ ace->e_type = RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+ ace->e_flags = RICHACE_SPECIAL_WHO;
+ ace->e_mask = allowed;
+ ace->e_id.special = RICHACE_EVERYONE_SPECIAL_ID;
+ }
+ }
+ return 0;
+}
--
2.4.3


2015-09-27 22:11:41

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 20/41] richacl: acl editing helper functions

The file masks in richacls make chmod and creating new files more
efficient than having to apply file permission bits to the acl directly.
They also allow us to regain permissions from an acl even after a
restrictive chmod, because the permissions in the acl itself are not
being destroyed. In POSIX ACLs, the mask entry has a similar function.

Protocols like nfsv4 do not understand file masks. For those protocols,
we need to compute nfs4 acls which represent the effective permissions
granted by a richacl: we need to "apply" the file masks to the acl.

This is the first in a series of richacl transformation patches; it
implements basic richacl editing functions. The following patches
implement algorithms for transforming a richacl so that it can be
evaluated as a plain nfs4 acl, with identical permission check results.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/Makefile | 3 +-
fs/richacl_compat.c | 155 +++++++++++++++++++++++++++++++++++++++++
include/linux/richacl_compat.h | 40 +++++++++++
3 files changed, 197 insertions(+), 1 deletion(-)
create mode 100644 fs/richacl_compat.c
create mode 100644 include/linux/richacl_compat.h

diff --git a/fs/Makefile b/fs/Makefile
index 35e640d..32b391b 100644
--- a/fs/Makefile
+++ b/fs/Makefile
@@ -49,7 +49,8 @@ obj-$(CONFIG_SYSCTL) += drop_caches.o

obj-$(CONFIG_FHANDLE) += fhandle.o
obj-$(CONFIG_FS_RICHACL) += richacl.o
-richacl-y := richacl_base.o richacl_inode.o richacl_xattr.o
+richacl-y := richacl_base.o richacl_inode.o \
+ richacl_xattr.o richacl_compat.o

obj-y += quota/

diff --git a/fs/richacl_compat.c b/fs/richacl_compat.c
new file mode 100644
index 0000000..341e429
--- /dev/null
+++ b/fs/richacl_compat.c
@@ -0,0 +1,155 @@
+/*
+ * Copyright (C) 2006, 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/richacl_compat.h>
+
+/**
+ * richacl_prepare - allocate richacl being constructed
+ *
+ * Allocate a richacl which can hold @count entries but which is initially
+ * empty.
+ */
+struct richacl *richacl_prepare(struct richacl_alloc *alloc, unsigned int count)
+{
+ alloc->acl = richacl_alloc(count, GFP_KERNEL);
+ if (!alloc->acl)
+ return NULL;
+ alloc->acl->a_count = 0;
+ alloc->count = count;
+ return alloc->acl;
+}
+EXPORT_SYMBOL_GPL(richacl_prepare);
+
+/**
+ * richacl_delete_entry - delete an entry in an acl
+ * @alloc: acl and number of allocated entries
+ * @ace: an entry in @alloc->acl
+ *
+ * Updates @ace so that it points to the entry before the deleted entry
+ * on return. (When deleting the first entry, @ace will point to the
+ * (non-existent) entry before the first entry). This behavior is the
+ * expected behavior when deleting entries while forward iterating over
+ * an acl.
+ */
+void
+richacl_delete_entry(struct richacl_alloc *alloc, struct richace **ace)
+{
+ void *end = alloc->acl->a_entries + alloc->acl->a_count;
+
+ memmove(*ace, *ace + 1, end - (void *)(*ace + 1));
+ (*ace)--;
+ alloc->acl->a_count--;
+}
+EXPORT_SYMBOL_GPL(richacl_delete_entry);
+
+/**
+ * richacl_insert_entry - insert an entry in an acl
+ * @alloc: acl and number of allocated entries
+ * @ace: entry before which the new entry shall be inserted
+ *
+ * Insert a new entry in @alloc->acl at position @ace and zero-initialize
+ * it. This may require reallocating @alloc->acl.
+ */
+int
+richacl_insert_entry(struct richacl_alloc *alloc, struct richace **ace)
+{
+ struct richacl *acl = alloc->acl;
+ unsigned int index = *ace - acl->a_entries;
+ size_t tail_size = (acl->a_count - index) * sizeof(struct richace);
+
+ if (alloc->count == acl->a_count) {
+ size_t new_size = sizeof(struct richacl) +
+ (acl->a_count + 1) * sizeof(struct richace);
+
+ acl = krealloc(acl, new_size, GFP_KERNEL);
+ if (!acl)
+ return -1;
+ *ace = acl->a_entries + index;
+ alloc->acl = acl;
+ alloc->count++;
+ }
+
+ memmove(*ace + 1, *ace, tail_size);
+ memset(*ace, 0, sizeof(**ace));
+ acl->a_count++;
+ return 0;
+}
+EXPORT_SYMBOL_GPL(richacl_insert_entry);
+
+/**
+ * richacl_append_entry - append an entry to an acl
+ * @alloc: acl and number of allocated entries
+ *
+ * This may require reallocating @alloc->acl.
+ */
+struct richace *richacl_append_entry(struct richacl_alloc *alloc)
+{
+ struct richacl *acl = alloc->acl;
+ struct richace *ace = acl->a_entries + acl->a_count;
+
+ if (alloc->count > alloc->acl->a_count) {
+ acl->a_count++;
+ return ace;
+ }
+ return richacl_insert_entry(alloc, &ace) ? NULL : ace;
+}
+EXPORT_SYMBOL_GPL(richacl_append_entry);
+
+/**
+ * richace_change_mask - set the mask of @ace to @mask
+ * @alloc: acl and number of allocated entries
+ * @ace: entry to modify
+ * @mask: new mask for @ace
+ *
+ * If @ace is inheritable, a inherit-only ace is inserted before @ace which
+ * includes the inheritable permissions of @ace and the inheritance flags of
+ * @ace are cleared before changing the mask.
+ *
+ * If @mask is 0, the original ace is turned into an inherit-only entry if
+ * there are any inheritable permissions, and removed otherwise.
+ *
+ * The returned @ace points to the modified or inserted effective-only acl
+ * entry if that entry exists, to the entry that has become inheritable-only,
+ * or else to the previous entry in the acl.
+ */
+static int
+richace_change_mask(struct richacl_alloc *alloc, struct richace **ace,
+ unsigned int mask)
+{
+ if (mask && (*ace)->e_mask == mask)
+ (*ace)->e_flags &= ~RICHACE_INHERIT_ONLY_ACE;
+ else if (mask & ~RICHACE_POSIX_ALWAYS_ALLOWED) {
+ if (richace_is_inheritable(*ace)) {
+ if (richacl_insert_entry(alloc, ace))
+ return -1;
+ richace_copy(*ace, *ace + 1);
+ (*ace)->e_flags |= RICHACE_INHERIT_ONLY_ACE;
+ (*ace)++;
+ (*ace)->e_flags &= ~RICHACE_INHERITANCE_FLAGS |
+ RICHACE_INHERITED_ACE;
+ }
+ (*ace)->e_mask = mask;
+ } else {
+ if (richace_is_inheritable(*ace))
+ (*ace)->e_flags |= RICHACE_INHERIT_ONLY_ACE;
+ else
+ richacl_delete_entry(alloc, ace);
+ }
+ return 0;
+}
diff --git a/include/linux/richacl_compat.h b/include/linux/richacl_compat.h
new file mode 100644
index 0000000..a9ff630
--- /dev/null
+++ b/include/linux/richacl_compat.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef __RICHACL_COMPAT_H
+#define __RICHACL_COMPAT_H
+
+#include <linux/richacl.h>
+
+/**
+ * struct richacl_alloc - remember how many entries are actually allocated
+ * @acl: acl with a_count <= @count
+ * @count: the actual number of entries allocated in @acl
+ *
+ * We pass around this structure while modifying an acl so that we do
+ * not have to reallocate when we remove existing entries followed by
+ * adding new entries.
+ */
+struct richacl_alloc {
+ struct richacl *acl;
+ unsigned int count;
+};
+
+struct richacl *richacl_prepare(struct richacl_alloc *, unsigned int);
+struct richace *richacl_append_entry(struct richacl_alloc *);
+int richacl_insert_entry(struct richacl_alloc *, struct richace **);
+void richacl_delete_entry(struct richacl_alloc *, struct richace **);
+
+#endif /* __RICHACL_COMPAT_H */
--
2.4.3


2015-09-27 22:11:35

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 19/41] ext4: Add richacl feature flag

From: "Aneesh Kumar K.V" <[email protected]>

This feature flag selects richacl instead of posix acl support on the
file system. In addition, the "acl" mount option is needed for enabling
either of the two kinds of acls.

Signed-off-by: Aneesh Kumar K.V <[email protected]>
Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/ext4/ext4.h | 6 ++++--
fs/ext4/super.c | 42 +++++++++++++++++++++++++++++++++---------
2 files changed, 37 insertions(+), 11 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index fd1f28b..b97a3b1 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -991,7 +991,7 @@ struct ext4_inode_info {
#define EXT4_MOUNT_UPDATE_JOURNAL 0x01000 /* Update the journal format */
#define EXT4_MOUNT_NO_UID32 0x02000 /* Disable 32-bit UIDs */
#define EXT4_MOUNT_XATTR_USER 0x04000 /* Extended user attributes */
-#define EXT4_MOUNT_POSIX_ACL 0x08000 /* POSIX Access Control Lists */
+#define EXT4_MOUNT_ACL 0x08000 /* Access Control Lists */
#define EXT4_MOUNT_NO_AUTO_DA_ALLOC 0x10000 /* No auto delalloc mapping */
#define EXT4_MOUNT_BARRIER 0x20000 /* Use block barriers */
#define EXT4_MOUNT_QUOTA 0x80000 /* Some quota option set */
@@ -1582,6 +1582,7 @@ static inline int ext4_encrypted_inode(struct inode *inode)
#define EXT4_FEATURE_INCOMPAT_LARGEDIR 0x4000 /* >2GB or 3-lvl htree */
#define EXT4_FEATURE_INCOMPAT_INLINE_DATA 0x8000 /* data in inode */
#define EXT4_FEATURE_INCOMPAT_ENCRYPT 0x10000
+#define EXT4_FEATURE_INCOMPAT_RICHACL 0x20000

#define EXT2_FEATURE_COMPAT_SUPP EXT4_FEATURE_COMPAT_EXT_ATTR
#define EXT2_FEATURE_INCOMPAT_SUPP (EXT4_FEATURE_INCOMPAT_FILETYPE| \
@@ -1607,7 +1608,8 @@ static inline int ext4_encrypted_inode(struct inode *inode)
EXT4_FEATURE_INCOMPAT_FLEX_BG| \
EXT4_FEATURE_INCOMPAT_MMP | \
EXT4_FEATURE_INCOMPAT_INLINE_DATA | \
- EXT4_FEATURE_INCOMPAT_ENCRYPT)
+ EXT4_FEATURE_INCOMPAT_ENCRYPT | \
+ EXT4_FEATURE_INCOMPAT_RICHACL)
#define EXT4_FEATURE_RO_COMPAT_SUPP (EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER| \
EXT4_FEATURE_RO_COMPAT_LARGE_FILE| \
EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index a63c7b0..84d1a5d 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1270,6 +1270,28 @@ static ext4_fsblk_t get_sb_block(void **data)
return sb_block;
}

+static int enable_acl(struct super_block *sb)
+{
+ sb->s_flags &= ~(MS_POSIXACL | MS_RICHACL);
+ if (test_opt(sb, ACL)) {
+ if (EXT4_HAS_INCOMPAT_FEATURE(sb,
+ EXT4_FEATURE_INCOMPAT_RICHACL)) {
+#ifdef CONFIG_EXT4_FS_RICHACL
+ sb->s_flags |= MS_RICHACL;
+#else
+ return -EOPNOTSUPP;
+#endif
+ } else {
+#ifdef CONFIG_EXT4_FS_POSIX_ACL
+ sb->s_flags |= MS_POSIXACL;
+#else
+ return -EOPNOTSUPP;
+#endif
+ }
+ }
+ return 0;
+}
+
#define DEFAULT_JOURNAL_IOPRIO (IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, 3))
static char deprecated_msg[] = "Mount option \"%s\" will be removed by %s\n"
"Contact [email protected] if you think we should keep it.\n";
@@ -1416,9 +1438,9 @@ static const struct mount_opts {
MOPT_NO_EXT2 | MOPT_DATAJ},
{Opt_user_xattr, EXT4_MOUNT_XATTR_USER, MOPT_SET},
{Opt_nouser_xattr, EXT4_MOUNT_XATTR_USER, MOPT_CLEAR},
-#ifdef CONFIG_EXT4_FS_POSIX_ACL
- {Opt_acl, EXT4_MOUNT_POSIX_ACL, MOPT_SET},
- {Opt_noacl, EXT4_MOUNT_POSIX_ACL, MOPT_CLEAR},
+#if defined(CONFIG_EXT4_FS_POSIX_ACL) || defined(CONFIG_EXT4_FS_RICHACL)
+ {Opt_acl, EXT4_MOUNT_ACL, MOPT_SET},
+ {Opt_noacl, EXT4_MOUNT_ACL, MOPT_CLEAR},
#else
{Opt_acl, 0, MOPT_NOSUPPORT},
{Opt_noacl, 0, MOPT_NOSUPPORT},
@@ -3576,8 +3598,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
set_opt(sb, NO_UID32);
/* xattr user namespace & acls are now defaulted on */
set_opt(sb, XATTR_USER);
-#ifdef CONFIG_EXT4_FS_POSIX_ACL
- set_opt(sb, POSIX_ACL);
+#if defined(CONFIG_EXT4_FS_POSIX_ACL) || defined(CONFIG_EXT4_FS_RICHACL)
+ set_opt(sb, ACL);
#endif
/* don't forget to enable journal_csum when metadata_csum is enabled. */
if (ext4_has_metadata_csum(sb))
@@ -3660,8 +3682,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
sb->s_iflags |= SB_I_CGROUPWB;
}

- sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
- (test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
+ err = enable_acl(sb);
+ if (err)
+ goto failed_mount;

if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV &&
(EXT4_HAS_COMPAT_FEATURE(sb, ~0U) ||
@@ -4981,8 +5004,9 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED)
ext4_abort(sb, "Abort forced by user");

- sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
- (test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0);
+ err = enable_acl(sb);
+ if (err)
+ goto restore_opts;

es = sbi->s_es;

--
2.4.3


2015-09-27 22:11:29

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 18/41] ext4: Add richacl support

From: "Aneesh Kumar K.V" <[email protected]>

Support the richacl permission model in ext4. The richacls are stored
in "system.richacl" xattrs. Richacls need to be enabled by tune2fs or
at file system create time.

Signed-off-by: Aneesh Kumar K.V <[email protected]>
Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/ext4/Kconfig | 15 ++++
fs/ext4/Makefile | 1 +
fs/ext4/acl.c | 6 +-
fs/ext4/acl.h | 12 +--
fs/ext4/file.c | 6 +-
fs/ext4/ialloc.c | 7 +-
fs/ext4/inode.c | 10 ++-
fs/ext4/namei.c | 11 ++-
fs/ext4/richacl.c | 213 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
fs/ext4/richacl.h | 47 ++++++++++++
fs/ext4/xattr.c | 6 ++
fs/ext4/xattr.h | 1 +
12 files changed, 316 insertions(+), 19 deletions(-)
create mode 100644 fs/ext4/richacl.c
create mode 100644 fs/ext4/richacl.h

diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig
index 47728da..5f8360a 100644
--- a/fs/ext4/Kconfig
+++ b/fs/ext4/Kconfig
@@ -130,3 +130,18 @@ config EXT4_DEBUG
If you select Y here, then you will be able to turn on debugging
with a command such as:
echo 1 > /sys/module/ext4/parameters/mballoc_debug
+
+config EXT4_FS_RICHACL
+ bool "Ext4 Rich Access Control Lists (EXPERIMENTAL)"
+ depends on EXT4_FS
+ select FS_RICHACL
+ help
+ Rich ACLs are an implementation of NFSv4 ACLs, extended by file masks
+ to fit into the standard POSIX file permission model. They are
+ designed to work seamlessly locally as well as across the NFSv4 and
+ CIFS/SMB2 network file system protocols.
+
+ To learn more about Rich ACL, visit
+ http://acl.bestbits.at/richacl/
+
+ If you don't know what Rich ACLs are, say N
diff --git a/fs/ext4/Makefile b/fs/ext4/Makefile
index 75285ea..ea0d539 100644
--- a/fs/ext4/Makefile
+++ b/fs/ext4/Makefile
@@ -14,3 +14,4 @@ ext4-$(CONFIG_EXT4_FS_POSIX_ACL) += acl.o
ext4-$(CONFIG_EXT4_FS_SECURITY) += xattr_security.o
ext4-$(CONFIG_EXT4_FS_ENCRYPTION) += crypto_policy.o crypto.o \
crypto_key.o crypto_fname.o
+ext4-$(CONFIG_EXT4_FS_RICHACL) += richacl.o
diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c
index 69b1e73..d965fa6 100644
--- a/fs/ext4/acl.c
+++ b/fs/ext4/acl.c
@@ -140,7 +140,7 @@ fail:
* inode->i_mutex: don't care
*/
struct posix_acl *
-ext4_get_acl(struct inode *inode, int type)
+ext4_get_posix_acl(struct inode *inode, int type)
{
int name_index;
char *value = NULL;
@@ -234,7 +234,7 @@ __ext4_set_acl(handle_t *handle, struct inode *inode, int type,
}

int
-ext4_set_acl(struct inode *inode, struct posix_acl *acl, int type)
+ext4_set_posix_acl(struct inode *inode, struct posix_acl *acl, int type)
{
handle_t *handle;
int error, retries = 0;
@@ -259,7 +259,7 @@ retry:
* inode->i_mutex: up (access to inode is still exclusive)
*/
int
-ext4_init_acl(handle_t *handle, struct inode *inode, struct inode *dir)
+ext4_init_posix_acl(handle_t *handle, struct inode *inode, struct inode *dir)
{
struct posix_acl *default_acl, *acl;
int error;
diff --git a/fs/ext4/acl.h b/fs/ext4/acl.h
index da2c795..450b4d1 100644
--- a/fs/ext4/acl.h
+++ b/fs/ext4/acl.h
@@ -54,17 +54,17 @@ static inline int ext4_acl_count(size_t size)
#ifdef CONFIG_EXT4_FS_POSIX_ACL

/* acl.c */
-struct posix_acl *ext4_get_acl(struct inode *inode, int type);
-int ext4_set_acl(struct inode *inode, struct posix_acl *acl, int type);
-extern int ext4_init_acl(handle_t *, struct inode *, struct inode *);
+struct posix_acl *ext4_get_posix_acl(struct inode *inode, int type);
+int ext4_set_posix_acl(struct inode *inode, struct posix_acl *acl, int type);
+extern int ext4_init_posix_acl(handle_t *, struct inode *, struct inode *);

#else /* CONFIG_EXT4_FS_POSIX_ACL */
#include <linux/sched.h>
-#define ext4_get_acl NULL
-#define ext4_set_acl NULL
+#define ext4_get_posix_acl NULL
+#define ext4_set_posix_acl NULL

static inline int
-ext4_init_acl(handle_t *handle, struct inode *inode, struct inode *dir)
+ext4_init_posix_acl(handle_t *handle, struct inode *inode, struct inode *dir)
{
return 0;
}
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 113837e..48818b2 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -30,6 +30,7 @@
#include "ext4_jbd2.h"
#include "xattr.h"
#include "acl.h"
+#include "richacl.h"

/*
* Called when an inode is released. Note that this is different
@@ -717,8 +718,9 @@ const struct inode_operations ext4_file_inode_operations = {
.getxattr = generic_getxattr,
.listxattr = ext4_listxattr,
.removexattr = generic_removexattr,
- .get_acl = ext4_get_acl,
- .set_acl = ext4_set_acl,
+ .get_acl = ext4_get_posix_acl,
+ .set_acl = ext4_set_posix_acl,
+ .get_richacl = ext4_get_richacl,
.fiemap = ext4_fiemap,
};

diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 619bfc1..5e701b3 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -27,6 +27,7 @@
#include "ext4_jbd2.h"
#include "xattr.h"
#include "acl.h"
+#include "richacl.h"

#include <trace/events/ext4.h>

@@ -1052,7 +1053,11 @@ got:
if (err)
goto fail_drop;

- err = ext4_init_acl(handle, inode, dir);
+ if (EXT4_IS_RICHACL(dir))
+ err = ext4_init_richacl(handle, inode, dir);
+ else
+ err = ext4_init_posix_acl(handle, inode, dir);
+
if (err)
goto fail_free_drop;

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 612fbcf..bdd1ad5 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -42,6 +42,7 @@
#include "xattr.h"
#include "acl.h"
#include "truncate.h"
+#include "richacl.h"

#include <trace/events/ext4.h>

@@ -4805,9 +4806,12 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr)
if (orphan && inode->i_nlink)
ext4_orphan_del(NULL, inode);

- if (!rc && (ia_valid & ATTR_MODE))
- rc = posix_acl_chmod(inode, inode->i_mode);
-
+ if (!rc && (ia_valid & ATTR_MODE)) {
+ if (EXT4_IS_RICHACL(inode))
+ rc = ext4_richacl_chmod(inode);
+ else
+ rc = posix_acl_chmod(inode, inode->i_mode);
+ }
err_out:
ext4_std_error(inode->i_sb, error);
if (!error)
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 9f61e76..83a492d 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -38,6 +38,7 @@

#include "xattr.h"
#include "acl.h"
+#include "richacl.h"

#include <trace/events/ext4.h>
/*
@@ -3852,8 +3853,9 @@ const struct inode_operations ext4_dir_inode_operations = {
.getxattr = generic_getxattr,
.listxattr = ext4_listxattr,
.removexattr = generic_removexattr,
- .get_acl = ext4_get_acl,
- .set_acl = ext4_set_acl,
+ .get_acl = ext4_get_posix_acl,
+ .set_acl = ext4_set_posix_acl,
+ .get_richacl = ext4_get_richacl,
.fiemap = ext4_fiemap,
};

@@ -3863,6 +3865,7 @@ const struct inode_operations ext4_special_inode_operations = {
.getxattr = generic_getxattr,
.listxattr = ext4_listxattr,
.removexattr = generic_removexattr,
- .get_acl = ext4_get_acl,
- .set_acl = ext4_set_acl,
+ .get_acl = ext4_get_posix_acl,
+ .set_acl = ext4_set_posix_acl,
+ .get_richacl = ext4_get_richacl,
};
diff --git a/fs/ext4/richacl.c b/fs/ext4/richacl.c
new file mode 100644
index 0000000..6758def
--- /dev/null
+++ b/fs/ext4/richacl.c
@@ -0,0 +1,213 @@
+/*
+ * Copyright IBM Corporation, 2010
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Author Aneesh Kumar K.V <[email protected]>
+ *
+ * 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/kernel.h>
+#include <linux/fs.h>
+#include <linux/richacl_xattr.h>
+
+#include "ext4.h"
+#include "ext4_jbd2.h"
+#include "xattr.h"
+#include "acl.h"
+#include "richacl.h"
+
+struct richacl *
+ext4_get_richacl(struct inode *inode)
+{
+ const int name_index = EXT4_XATTR_INDEX_RICHACL;
+ void *value = NULL;
+ struct richacl *acl;
+ int retval;
+
+ if (!IS_RICHACL(inode))
+ return ERR_PTR(-EOPNOTSUPP);
+ acl = get_cached_richacl(inode);
+ if (acl != ACL_NOT_CACHED)
+ return acl;
+ retval = ext4_xattr_get(inode, name_index, "", NULL, 0);
+ if (retval > 0) {
+ value = kmalloc(retval, GFP_NOFS);
+ if (!value)
+ return ERR_PTR(-ENOMEM);
+ retval = ext4_xattr_get(inode, name_index, "", value, retval);
+ }
+ if (retval > 0) {
+ acl = richacl_from_xattr(&init_user_ns, value, retval);
+ if (acl == ERR_PTR(-EINVAL))
+ acl = ERR_PTR(-EIO);
+ } else if (retval == -ENODATA || retval == -ENOSYS)
+ acl = NULL;
+ else
+ acl = ERR_PTR(retval);
+ kfree(value);
+
+ if (!IS_ERR(acl))
+ set_cached_richacl(inode, acl);
+
+ return acl;
+}
+
+static int
+ext4_set_richacl(handle_t *handle, struct inode *inode, struct richacl *acl)
+{
+ const int name_index = EXT4_XATTR_INDEX_RICHACL;
+ size_t size = 0;
+ void *value = NULL;
+ int retval;
+
+ if (acl) {
+ mode_t mode = inode->i_mode;
+
+ if (richacl_equiv_mode(acl, &mode) == 0) {
+ inode->i_mode = mode;
+ ext4_mark_inode_dirty(handle, inode);
+ acl = NULL;
+ }
+ }
+ if (acl) {
+ size = richacl_xattr_size(acl);
+ value = kmalloc(size, GFP_NOFS);
+ if (!value)
+ return -ENOMEM;
+ richacl_to_xattr(&init_user_ns, acl, value, size);
+ }
+ if (handle)
+ retval = ext4_xattr_set_handle(handle, inode, name_index, "",
+ value, size, 0);
+ else
+ retval = ext4_xattr_set(inode, name_index, "", value, size, 0);
+ kfree(value);
+ if (!retval)
+ set_cached_richacl(inode, acl);
+
+ return retval;
+}
+
+int
+ext4_init_richacl(handle_t *handle, struct inode *inode, struct inode *dir)
+{
+ struct richacl *acl = richacl_create(inode, dir);
+ int error;
+
+ error = PTR_ERR(acl);
+ if (IS_ERR(acl))
+ return error;
+ if (acl) {
+ error = ext4_set_richacl(handle, inode, acl);
+ richacl_put(acl);
+ }
+ return error;
+}
+
+int
+ext4_richacl_chmod(struct inode *inode)
+{
+ struct richacl *acl;
+ int retval;
+
+ if (S_ISLNK(inode->i_mode))
+ return -EOPNOTSUPP;
+ acl = ext4_get_richacl(inode);
+ if (IS_ERR_OR_NULL(acl))
+ return PTR_ERR(acl);
+ acl = richacl_chmod(acl, inode->i_mode);
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+ retval = ext4_set_richacl(NULL, inode, acl);
+ richacl_put(acl);
+
+ return retval;
+}
+
+static size_t
+ext4_xattr_list_richacl(struct dentry *dentry, char *list, size_t list_len,
+ const char *name, size_t name_len, int type)
+{
+ const size_t size = sizeof(XATTR_NAME_RICHACL);
+
+ if (!IS_RICHACL(d_inode(dentry)))
+ return 0;
+ if (list && size <= list_len)
+ memcpy(list, XATTR_NAME_RICHACL, size);
+ return size;
+}
+
+static int
+ext4_xattr_get_richacl(struct dentry *dentry, const char *name, void *buffer,
+ size_t buffer_size, int type)
+{
+ struct richacl *acl;
+ int error;
+
+ if (strcmp(name, "") != 0)
+ return -EINVAL;
+ acl = ext4_get_richacl(d_inode(dentry));
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+ if (acl == NULL)
+ return -ENODATA;
+
+ error = richacl_to_xattr(&init_user_ns, acl, buffer, buffer_size);
+ richacl_put(acl);
+ return error;
+}
+
+static int
+ext4_xattr_set_richacl(struct dentry *dentry, const char *name,
+ const void *value, size_t size, int flags, int type)
+{
+ handle_t *handle;
+ struct richacl *acl = NULL;
+ int retval, retries = 0;
+ struct inode *inode = d_inode(dentry);
+
+ if (!IS_RICHACL(d_inode(dentry)))
+ return -EOPNOTSUPP;
+ if (S_ISLNK(inode->i_mode))
+ return -EOPNOTSUPP;
+ if (strcmp(name, "") != 0)
+ return -EINVAL;
+ if (!uid_eq(current_fsuid(), inode->i_uid) &&
+ inode_permission(inode, MAY_CHMOD) &&
+ !capable(CAP_FOWNER))
+ return -EPERM;
+ if (value) {
+ acl = richacl_from_xattr(&init_user_ns, value, size);
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+
+ inode->i_mode &= ~S_IRWXUGO;
+ inode->i_mode |= richacl_masks_to_mode(acl);
+ }
+
+retry:
+ handle = ext4_journal_start(inode, EXT4_HT_XATTR,
+ EXT4_DATA_TRANS_BLOCKS(inode->i_sb));
+ if (IS_ERR(handle))
+ return PTR_ERR(handle);
+ retval = ext4_set_richacl(handle, inode, acl);
+ ext4_journal_stop(handle);
+ if (retval == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
+ goto retry;
+ richacl_put(acl);
+ return retval;
+}
+
+const struct xattr_handler ext4_richacl_xattr_handler = {
+ .prefix = XATTR_NAME_RICHACL,
+ .list = ext4_xattr_list_richacl,
+ .get = ext4_xattr_get_richacl,
+ .set = ext4_xattr_set_richacl,
+};
diff --git a/fs/ext4/richacl.h b/fs/ext4/richacl.h
new file mode 100644
index 0000000..09a5cad
--- /dev/null
+++ b/fs/ext4/richacl.h
@@ -0,0 +1,47 @@
+/*
+ * Copyright IBM Corporation, 2010
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Author Aneesh Kumar K.V <[email protected]>
+ *
+ * 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_EXT4_RICHACL_H
+#define __FS_EXT4_RICHACL_H
+
+#include <linux/richacl.h>
+
+#ifdef CONFIG_EXT4_FS_RICHACL
+
+#define EXT4_IS_RICHACL(inode) IS_RICHACL(inode)
+
+extern struct richacl *ext4_get_richacl(struct inode *);
+extern int ext4_init_richacl(handle_t *, struct inode *, struct inode *);
+extern int ext4_richacl_chmod(struct inode *);
+
+#else /* CONFIG_FS_EXT4_RICHACL */
+
+#define EXT4_IS_RICHACL(inode) (0)
+#define ext4_get_richacl NULL
+
+static inline int
+ext4_init_richacl(handle_t *handle, struct inode *inode, struct inode *dir)
+{
+ return 0;
+}
+
+static inline int
+ext4_richacl_chmod(struct inode *inode)
+{
+ return 0;
+}
+
+#endif /* CONFIG_FS_EXT4_RICHACL */
+#endif /* __FS_EXT4_RICHACL_H */
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 16e28c0..c1dee9b 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -99,6 +99,9 @@ static const struct xattr_handler *ext4_xattr_handler_map[] = {
#ifdef CONFIG_EXT4_FS_SECURITY
[EXT4_XATTR_INDEX_SECURITY] = &ext4_xattr_security_handler,
#endif
+#ifdef CONFIG_EXT4_FS_RICHACL
+ [EXT4_XATTR_INDEX_RICHACL] = &ext4_richacl_xattr_handler,
+#endif
};

const struct xattr_handler *ext4_xattr_handlers[] = {
@@ -111,6 +114,9 @@ const struct xattr_handler *ext4_xattr_handlers[] = {
#ifdef CONFIG_EXT4_FS_SECURITY
&ext4_xattr_security_handler,
#endif
+#ifdef CONFIG_EXT4_FS_RICHACL
+ &ext4_richacl_xattr_handler,
+#endif
NULL
};

diff --git a/fs/ext4/xattr.h b/fs/ext4/xattr.h
index ddc0957..f315493 100644
--- a/fs/ext4/xattr.h
+++ b/fs/ext4/xattr.h
@@ -98,6 +98,7 @@ struct ext4_xattr_ibody_find {
extern const struct xattr_handler ext4_xattr_user_handler;
extern const struct xattr_handler ext4_xattr_trusted_handler;
extern const struct xattr_handler ext4_xattr_security_handler;
+extern const struct xattr_handler ext4_richacl_xattr_handler;

#define EXT4_XATTR_NAME_ENCRYPTION_CONTEXT "c"

--
2.4.3


2015-09-27 22:11:23

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 17/41] vfs: Add richacl permission checking

Hook the richacl permission checking function into the vfs.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/namei.c | 51 +++++++++++++++++++++++++++++++++++++++++++++++++--
fs/posix_acl.c | 6 +++---
2 files changed, 52 insertions(+), 5 deletions(-)

diff --git a/fs/namei.c b/fs/namei.c
index 6865cdf..3cc13f0 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -35,6 +35,7 @@
#include <linux/fs_struct.h>
#include <linux/posix_acl.h>
#include <linux/hash.h>
+#include <linux/richacl.h>
#include <asm/uaccess.h>

#include "internal.h"
@@ -255,7 +256,40 @@ void putname(struct filename *name)
__putname(name);
}

-static int check_acl(struct inode *inode, int mask)
+static int check_richacl(struct inode *inode, int mask)
+{
+#ifdef CONFIG_FS_RICHACL
+ struct richacl *acl;
+
+ if (mask & MAY_NOT_BLOCK) {
+ acl = get_cached_richacl_rcu(inode);
+ if (!acl)
+ goto no_acl;
+ /* no ->get_richacl() calls in RCU mode... */
+ if (acl == ACL_NOT_CACHED)
+ return -ECHILD;
+ return richacl_permission(inode, acl, mask & ~MAY_NOT_BLOCK);
+ }
+
+ acl = get_richacl(inode);
+ if (IS_ERR(acl))
+ return PTR_ERR(acl);
+ if (acl) {
+ int error = richacl_permission(inode, acl, mask);
+ richacl_put(acl);
+ return error;
+ }
+no_acl:
+#endif
+ if (mask & (MAY_DELETE_SELF | MAY_TAKE_OWNERSHIP |
+ MAY_CHMOD | MAY_SET_TIMES)) {
+ /* File permission bits cannot grant this. */
+ return -EACCES;
+ }
+ return -EAGAIN;
+}
+
+static int check_posix_acl(struct inode *inode, int mask)
{
#ifdef CONFIG_FS_POSIX_ACL
struct posix_acl *acl;
@@ -290,11 +324,24 @@ static int acl_permission_check(struct inode *inode, int mask)
{
unsigned int mode = inode->i_mode;

+ /*
+ * With POSIX ACLs, the (mode & S_IRWXU) bits exactly match the owner
+ * permissions, and we can skip checking posix acls for the owner.
+ * With richacls, the owner may be granted fewer permissions than the
+ * mode bits seem to suggest (for example, append but not write), and
+ * we always need to check the richacl.
+ */
+
+ if (IS_RICHACL(inode)) {
+ int error = check_richacl(inode, mask);
+ if (error != -EAGAIN)
+ return error;
+ }
if (likely(uid_eq(current_fsuid(), inode->i_uid)))
mode >>= 6;
else {
if (IS_POSIXACL(inode) && (mode & S_IRWXG)) {
- int error = check_acl(inode, mask);
+ int error = check_posix_acl(inode, mask);
if (error != -EAGAIN)
return error;
}
diff --git a/fs/posix_acl.c b/fs/posix_acl.c
index 1d766a5..3459bd5 100644
--- a/fs/posix_acl.c
+++ b/fs/posix_acl.c
@@ -100,13 +100,13 @@ struct posix_acl *get_acl(struct inode *inode, int type)
{
struct posix_acl *acl;

+ if (!IS_POSIXACL(inode))
+ return NULL;
+
acl = get_cached_acl(inode, type);
if (acl != ACL_NOT_CACHED)
return acl;

- if (!IS_POSIXACL(inode))
- return NULL;
-
/*
* A filesystem can force a ACL callback by just never filling the
* ACL cache. But normally you'd fill the cache either at inode
--
2.4.3


2015-09-27 22:11:17

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 16/41] richacl: xattr mapping functions

Map between "system.richacl" xattrs and the in-kernel representation.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/Makefile | 2 +-
fs/richacl_xattr.c | 220 ++++++++++++++++++++++++++++++++++++++++++
fs/xattr.c | 34 +++++--
include/linux/richacl_xattr.h | 62 ++++++++++++
include/uapi/linux/xattr.h | 2 +
5 files changed, 313 insertions(+), 7 deletions(-)
create mode 100644 fs/richacl_xattr.c
create mode 100644 include/linux/richacl_xattr.h

diff --git a/fs/Makefile b/fs/Makefile
index ec665fd..35e640d 100644
--- a/fs/Makefile
+++ b/fs/Makefile
@@ -49,7 +49,7 @@ obj-$(CONFIG_SYSCTL) += drop_caches.o

obj-$(CONFIG_FHANDLE) += fhandle.o
obj-$(CONFIG_FS_RICHACL) += richacl.o
-richacl-y := richacl_base.o richacl_inode.o
+richacl-y := richacl_base.o richacl_inode.o richacl_xattr.o

obj-y += quota/

diff --git a/fs/richacl_xattr.c b/fs/richacl_xattr.c
new file mode 100644
index 0000000..cd9979d
--- /dev/null
+++ b/fs/richacl_xattr.c
@@ -0,0 +1,220 @@
+/*
+ * Copyright (C) 2006, 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/sched.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <linux/richacl_xattr.h>
+
+MODULE_LICENSE("GPL");
+
+/**
+ * richacl_from_xattr - convert a richacl xattr into the in-memory representation
+ */
+struct richacl *
+richacl_from_xattr(struct user_namespace *user_ns,
+ const void *value, size_t size)
+{
+ const struct richacl_xattr *xattr_acl = value;
+ const struct richace_xattr *xattr_ace = (void *)(xattr_acl + 1);
+ struct richacl *acl;
+ struct richace *ace;
+ int count;
+
+ if (size < sizeof(*xattr_acl) ||
+ xattr_acl->a_version != RICHACL_XATTR_VERSION ||
+ (xattr_acl->a_flags & ~RICHACL_VALID_FLAGS))
+ return ERR_PTR(-EINVAL);
+ size -= sizeof(*xattr_acl);
+ count = le16_to_cpu(xattr_acl->a_count);
+ if (count > RICHACL_XATTR_MAX_COUNT)
+ return ERR_PTR(-EINVAL);
+ if (size != count * sizeof(*xattr_ace))
+ return ERR_PTR(-EINVAL);
+
+ acl = richacl_alloc(count, GFP_NOFS);
+ if (!acl)
+ return ERR_PTR(-ENOMEM);
+
+ acl->a_flags = xattr_acl->a_flags;
+ acl->a_owner_mask = le32_to_cpu(xattr_acl->a_owner_mask);
+ if (acl->a_owner_mask & ~RICHACE_VALID_MASK)
+ goto fail_einval;
+ acl->a_group_mask = le32_to_cpu(xattr_acl->a_group_mask);
+ if (acl->a_group_mask & ~RICHACE_VALID_MASK)
+ goto fail_einval;
+ acl->a_other_mask = le32_to_cpu(xattr_acl->a_other_mask);
+ if (acl->a_other_mask & ~RICHACE_VALID_MASK)
+ goto fail_einval;
+
+ richacl_for_each_entry(ace, acl) {
+ ace->e_type = le16_to_cpu(xattr_ace->e_type);
+ ace->e_flags = le16_to_cpu(xattr_ace->e_flags);
+ ace->e_mask = le32_to_cpu(xattr_ace->e_mask);
+
+ if (ace->e_flags & ~RICHACE_VALID_FLAGS)
+ goto fail_einval;
+ if (ace->e_flags & RICHACE_SPECIAL_WHO) {
+ ace->e_id.special = le32_to_cpu(xattr_ace->e_id);
+ if (ace->e_id.special > RICHACE_EVERYONE_SPECIAL_ID)
+ goto fail_einval;
+ } else if (ace->e_flags & RICHACE_IDENTIFIER_GROUP) {
+ u32 id = le32_to_cpu(xattr_ace->e_id);
+
+ ace->e_id.gid = make_kgid(user_ns, id);
+ if (!gid_valid(ace->e_id.gid))
+ goto fail_einval;
+ } else {
+ u32 id = le32_to_cpu(xattr_ace->e_id);
+
+ ace->e_id.uid = make_kuid(user_ns, id);
+ if (!uid_valid(ace->e_id.uid))
+ goto fail_einval;
+ }
+ if (ace->e_type > RICHACE_ACCESS_DENIED_ACE_TYPE ||
+ (ace->e_mask & ~RICHACE_VALID_MASK))
+ goto fail_einval;
+
+ xattr_ace++;
+ }
+
+ return acl;
+
+fail_einval:
+ richacl_put(acl);
+ return ERR_PTR(-EINVAL);
+}
+EXPORT_SYMBOL_GPL(richacl_from_xattr);
+
+/**
+ * richacl_xattr_size - compute the size of the xattr representation of @acl
+ */
+size_t
+richacl_xattr_size(const struct richacl *acl)
+{
+ size_t size = sizeof(struct richacl_xattr);
+
+ size += sizeof(struct richace_xattr) * acl->a_count;
+ return size;
+}
+EXPORT_SYMBOL_GPL(richacl_xattr_size);
+
+/**
+ * richacl_to_xattr - convert @acl into its xattr representation
+ * @acl: the richacl to convert
+ * @buffer: buffer for the result
+ * @size: size of @buffer
+ */
+int
+richacl_to_xattr(struct user_namespace *user_ns,
+ const struct richacl *acl, void *buffer, size_t size)
+{
+ struct richacl_xattr *xattr_acl = buffer;
+ struct richace_xattr *xattr_ace;
+ const struct richace *ace;
+ size_t real_size;
+
+ real_size = richacl_xattr_size(acl);
+ if (!buffer)
+ return real_size;
+ if (real_size > size)
+ return -ERANGE;
+
+ xattr_acl->a_version = RICHACL_XATTR_VERSION;
+ xattr_acl->a_flags = acl->a_flags;
+ xattr_acl->a_count = cpu_to_le16(acl->a_count);
+
+ xattr_acl->a_owner_mask = cpu_to_le32(acl->a_owner_mask);
+ xattr_acl->a_group_mask = cpu_to_le32(acl->a_group_mask);
+ xattr_acl->a_other_mask = cpu_to_le32(acl->a_other_mask);
+
+ xattr_ace = (void *)(xattr_acl + 1);
+ richacl_for_each_entry(ace, acl) {
+ xattr_ace->e_type = cpu_to_le16(ace->e_type);
+ xattr_ace->e_flags = cpu_to_le16(ace->e_flags);
+ xattr_ace->e_mask = cpu_to_le32(ace->e_mask);
+ if (ace->e_flags & RICHACE_SPECIAL_WHO)
+ xattr_ace->e_id = cpu_to_le32(ace->e_id.special);
+ else if (ace->e_flags & RICHACE_IDENTIFIER_GROUP)
+ xattr_ace->e_id =
+ cpu_to_le32(from_kgid(user_ns, ace->e_id.gid));
+ else
+ xattr_ace->e_id =
+ cpu_to_le32(from_kuid(user_ns, ace->e_id.uid));
+ xattr_ace++;
+ }
+ return real_size;
+}
+EXPORT_SYMBOL_GPL(richacl_to_xattr);
+
+/*
+ * Fix up the uids and gids in richacl extended attributes in place.
+ */
+static void richacl_fix_xattr_userns(
+ struct user_namespace *to, struct user_namespace *from,
+ void *value, size_t size)
+{
+ struct richacl_xattr *xattr_acl = value;
+ struct richace_xattr *xattr_ace =
+ (struct richace_xattr *)(xattr_acl + 1);
+ unsigned int count;
+
+ if (!value)
+ return;
+ if (size < sizeof(*xattr_acl))
+ return;
+ if (xattr_acl->a_version != cpu_to_le32(RICHACL_XATTR_VERSION))
+ return;
+ size -= sizeof(*xattr_acl);
+ if (size % sizeof(*xattr_ace))
+ return;
+ count = size / sizeof(*xattr_ace);
+ for (; count; count--, xattr_ace++) {
+ if (xattr_ace->e_flags & cpu_to_le16(RICHACE_SPECIAL_WHO))
+ continue;
+ if (xattr_ace->e_flags &
+ cpu_to_le16(RICHACE_IDENTIFIER_GROUP)) {
+ u32 id = le32_to_cpu(xattr_ace->e_id);
+ kgid_t gid = make_kgid(from, id);
+
+ xattr_ace->e_id = cpu_to_le32(from_kgid(to, gid));
+ } else {
+ u32 id = le32_to_cpu(xattr_ace->e_id);
+ kuid_t uid = make_kuid(from, id);
+
+ xattr_ace->e_id = cpu_to_le32(from_kuid(to, uid));
+ }
+ }
+}
+
+void richacl_fix_xattr_from_user(void *value, size_t size)
+{
+ struct user_namespace *user_ns = current_user_ns();
+
+ if (user_ns == &init_user_ns)
+ return;
+ richacl_fix_xattr_userns(&init_user_ns, user_ns, value, size);
+}
+
+void richacl_fix_xattr_to_user(void *value, size_t size)
+{
+ struct user_namespace *user_ns = current_user_ns();
+
+ if (user_ns == &init_user_ns)
+ return;
+ richacl_fix_xattr_userns(user_ns, &init_user_ns, value, size);
+}
diff --git a/fs/xattr.c b/fs/xattr.c
index 072fee1..f2313c6 100644
--- a/fs/xattr.c
+++ b/fs/xattr.c
@@ -21,6 +21,7 @@
#include <linux/audit.h>
#include <linux/vmalloc.h>
#include <linux/posix_acl_xattr.h>
+#include <linux/richacl_xattr.h>

#include <asm/uaccess.h>

@@ -314,6 +315,18 @@ out:
}
EXPORT_SYMBOL_GPL(vfs_removexattr);

+static void
+fix_xattr_from_user(const char *kname, void *kvalue, size_t size)
+{
+ if (strncmp(kname, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
+ return;
+ kname += XATTR_SYSTEM_PREFIX_LEN;
+ if (!strcmp(kname, XATTR_POSIX_ACL_ACCESS) ||
+ !strcmp(kname, XATTR_POSIX_ACL_DEFAULT))
+ posix_acl_fix_xattr_from_user(kvalue, size);
+ else if (!strcmp(kname, XATTR_RICHACL))
+ richacl_fix_xattr_from_user(kvalue, size);
+}

/*
* Extended attribute SET operations
@@ -350,9 +363,7 @@ setxattr(struct dentry *d, const char __user *name, const void __user *value,
error = -EFAULT;
goto out;
}
- if ((strcmp(kname, XATTR_NAME_POSIX_ACL_ACCESS) == 0) ||
- (strcmp(kname, XATTR_NAME_POSIX_ACL_DEFAULT) == 0))
- posix_acl_fix_xattr_from_user(kvalue, size);
+ fix_xattr_from_user(kname, kvalue, size);
}

error = vfs_setxattr(d, kname, kvalue, size, flags);
@@ -419,6 +430,19 @@ SYSCALL_DEFINE5(fsetxattr, int, fd, const char __user *, name,
return error;
}

+static void
+fix_xattr_to_user(const char *kname, void *kvalue, size_t size)
+{
+ if (strncmp(kname, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
+ return;
+ kname += XATTR_SYSTEM_PREFIX_LEN;
+ if (!strcmp(kname, XATTR_POSIX_ACL_ACCESS) ||
+ !strcmp(kname, XATTR_POSIX_ACL_DEFAULT))
+ posix_acl_fix_xattr_to_user(kvalue, size);
+ else if (!strcmp(kname, XATTR_RICHACL))
+ richacl_fix_xattr_to_user(kvalue, size);
+}
+
/*
* Extended attribute GET operations
*/
@@ -451,9 +475,7 @@ getxattr(struct dentry *d, const char __user *name, void __user *value,

error = vfs_getxattr(d, kname, kvalue, size);
if (error > 0) {
- if ((strcmp(kname, XATTR_NAME_POSIX_ACL_ACCESS) == 0) ||
- (strcmp(kname, XATTR_NAME_POSIX_ACL_DEFAULT) == 0))
- posix_acl_fix_xattr_to_user(kvalue, size);
+ fix_xattr_to_user(kname, kvalue, size);
if (size && copy_to_user(value, kvalue, error))
error = -EFAULT;
} else if (error == -ERANGE && size >= XATTR_SIZE_MAX) {
diff --git a/include/linux/richacl_xattr.h b/include/linux/richacl_xattr.h
new file mode 100644
index 0000000..f84cc21
--- /dev/null
+++ b/include/linux/richacl_xattr.h
@@ -0,0 +1,62 @@
+/*
+ * Copyright (C) 2006, 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef __RICHACL_XATTR_H
+#define __RICHACL_XATTR_H
+
+#include <linux/richacl.h>
+
+struct richace_xattr {
+ __le16 e_type;
+ __le16 e_flags;
+ __le32 e_mask;
+ __le32 e_id;
+};
+
+struct richacl_xattr {
+ unsigned char a_version;
+ unsigned char a_flags;
+ __le16 a_count;
+ __le32 a_owner_mask;
+ __le32 a_group_mask;
+ __le32 a_other_mask;
+};
+
+#define RICHACL_XATTR_VERSION 0
+#define RICHACL_XATTR_MAX_COUNT \
+ ((XATTR_SIZE_MAX - sizeof(struct richacl_xattr)) / \
+ sizeof(struct richace_xattr))
+
+extern struct richacl *richacl_from_xattr(struct user_namespace *, const void *,
+ size_t);
+extern size_t richacl_xattr_size(const struct richacl *);
+extern int richacl_to_xattr(struct user_namespace *, const struct richacl *,
+ void *, size_t);
+
+#ifdef CONFIG_FS_RICHACL
+extern void richacl_fix_xattr_from_user(void *, size_t);
+extern void richacl_fix_xattr_to_user(void *, size_t);
+#else
+static inline void richacl_fix_xattr_from_user(void *value, size_t size)
+{
+}
+
+static inline void richacl_fix_xattr_to_user(void *value, size_t size)
+{
+}
+#endif
+
+#endif /* __RICHACL_XATTR_H */
diff --git a/include/uapi/linux/xattr.h b/include/uapi/linux/xattr.h
index 1590c49..1996903 100644
--- a/include/uapi/linux/xattr.h
+++ b/include/uapi/linux/xattr.h
@@ -73,5 +73,7 @@
#define XATTR_POSIX_ACL_DEFAULT "posix_acl_default"
#define XATTR_NAME_POSIX_ACL_DEFAULT XATTR_SYSTEM_PREFIX XATTR_POSIX_ACL_DEFAULT

+#define XATTR_RICHACL "richacl"
+#define XATTR_NAME_RICHACL XATTR_SYSTEM_PREFIX XATTR_RICHACL

#endif /* _UAPI_LINUX_XATTR_H */
--
2.4.3


2015-09-27 22:11:11

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 15/41] richacl: Automatic Inheritance

Automatic Inheritance (AI) allows changes to the acl of a directory to
propagate down to children.

This is mostly implemented in user space: when a process changes the
permissions of a directory and Automatic Inheritance is enabled for that
directory, the process must propagate those changes to all children,
recursively.

The kernel enables this by keeping track of which permissions have been
inherited at create time. In addition, it makes sure that permission
propagation is turned off when the permissions are set explicitly (for
example, upon create or chmod).

Automatic Inheritance works as follows:

- When the RICHACL_AUTO_INHERIT flag in the acl of a file or directory
is not set, the file or directory is not affected by AI.

- When the RICHACL_AUTO_INHERIT flag in the acl of a directory is set
and a file or subdirectory is created in that directory, the
inherited acl will have the RICHACL_AUTO_INHERIT flag set, and all
inherited aces will have the RICHACE_INHERITED_ACE flag set. This
allows user space to distinguish between aces which have been
inherited and aces which have been explicitly added.

- When the RICHACL_PROTECTED acl flag in the acl of a file or directory
is set, AI will not modify the acl. This does not affect propagation
of permissions from the file to its children (if the file is a
directory).

Linux does not have a way of creating files or directories without setting the
file permission bits, so all files created inside a directory with
RICHACL_AUTO_INHERIT set will have the RICHACL_PROTECTED flag set. This
effectively disables Automatic Inheritance.

Protocols which support creating files without specifying permissions can
explicitly clear the RICHACL_PROTECTED flag after creating a file and reset the
file masks to "undo" applying the create mode; see richacl_compute_max_masks().
They should set the RICHACL_DEFAULTED flag. (A mechanism that would allow to
indicate to the kernel to ignore the create mode in the first place when there
are inherited permissions would be nice to have.)

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_base.c | 10 +++++++++-
fs/richacl_inode.c | 7 +++++++
include/linux/richacl.h | 23 ++++++++++++++++++++++-
3 files changed, 38 insertions(+), 2 deletions(-)

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index c1490c7..e0357ae 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -363,7 +363,8 @@ richacl_chmod(struct richacl *acl, mode_t mode)
if (acl->a_owner_mask == owner_mask &&
acl->a_group_mask == group_mask &&
acl->a_other_mask == other_mask &&
- (acl->a_flags & (RICHACL_WRITE_THROUGH | RICHACL_MASKED)))
+ (acl->a_flags & (RICHACL_WRITE_THROUGH | RICHACL_MASKED)) &&
+ (!richacl_is_auto_inherit(acl) || richacl_is_protected(acl)))
return acl;

clone = richacl_clone(acl, GFP_KERNEL);
@@ -375,6 +376,8 @@ richacl_chmod(struct richacl *acl, mode_t mode)
clone->a_owner_mask = owner_mask;
clone->a_group_mask = group_mask;
clone->a_other_mask = other_mask;
+ if (richacl_is_auto_inherit(clone))
+ clone->a_flags |= RICHACL_PROTECTED;

return clone;
}
@@ -548,6 +551,11 @@ richacl_inherit(const struct richacl *dir_acl, int isdir)
ace++;
}
}
+ if (richacl_is_auto_inherit(dir_acl)) {
+ acl->a_flags = RICHACL_AUTO_INHERIT;
+ richacl_for_each_entry(ace, acl)
+ ace->e_flags |= RICHACE_INHERITED_ACE;
+ }

return acl;
}
diff --git a/fs/richacl_inode.c b/fs/richacl_inode.c
index 737f054..d451e83 100644
--- a/fs/richacl_inode.c
+++ b/fs/richacl_inode.c
@@ -248,6 +248,13 @@ richacl_inherit_inode(const struct richacl *dir_acl, struct inode *inode)
richacl_put(acl);
acl = NULL;
} else {
+ /*
+ * We need to set RICHACL_PROTECTED because we are
+ * doing an implicit chmod
+ */
+ if (richacl_is_auto_inherit(acl))
+ acl->a_flags |= RICHACL_PROTECTED;
+
richacl_compute_max_masks(acl);
/*
* Ensure that the acl will not grant any permissions
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index 9bf95c2..832b06c 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -53,10 +53,16 @@ struct richacl {
_ace--)

/* a_flags values */
+#define RICHACL_AUTO_INHERIT 0x01
+#define RICHACL_PROTECTED 0x02
+#define RICHACL_DEFAULTED 0x04
#define RICHACL_WRITE_THROUGH 0x40
#define RICHACL_MASKED 0x80

#define RICHACL_VALID_FLAGS ( \
+ RICHACL_AUTO_INHERIT | \
+ RICHACL_PROTECTED | \
+ RICHACL_DEFAULTED | \
RICHACL_WRITE_THROUGH | \
RICHACL_MASKED)

@@ -70,6 +76,7 @@ struct richacl {
#define RICHACE_NO_PROPAGATE_INHERIT_ACE 0x0004
#define RICHACE_INHERIT_ONLY_ACE 0x0008
#define RICHACE_IDENTIFIER_GROUP 0x0040
+#define RICHACE_INHERITED_ACE 0x0080
#define RICHACE_SPECIAL_WHO 0x4000

#define RICHACE_VALID_FLAGS ( \
@@ -78,13 +85,15 @@ struct richacl {
RICHACE_NO_PROPAGATE_INHERIT_ACE | \
RICHACE_INHERIT_ONLY_ACE | \
RICHACE_IDENTIFIER_GROUP | \
+ RICHACE_INHERITED_ACE | \
RICHACE_SPECIAL_WHO)

#define RICHACE_INHERITANCE_FLAGS ( \
RICHACE_FILE_INHERIT_ACE | \
RICHACE_DIRECTORY_INHERIT_ACE | \
RICHACE_NO_PROPAGATE_INHERIT_ACE | \
- RICHACE_INHERIT_ONLY_ACE )
+ RICHACE_INHERIT_ONLY_ACE | \
+ RICHACE_INHERITED_ACE )

/* e_mask bitflags */
#define RICHACE_READ_DATA 0x00000001
@@ -195,6 +204,18 @@ extern void set_cached_richacl(struct inode *, struct richacl *);
extern void forget_cached_richacl(struct inode *);
extern struct richacl *get_richacl(struct inode *);

+static inline int
+richacl_is_auto_inherit(const struct richacl *acl)
+{
+ return acl->a_flags & RICHACL_AUTO_INHERIT;
+}
+
+static inline int
+richacl_is_protected(const struct richacl *acl)
+{
+ return acl->a_flags & RICHACL_PROTECTED;
+}
+
/**
* richace_is_owner - check if @ace is an OWNER@ entry
*/
--
2.4.3


2015-09-27 22:11:06

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 14/41] richacl: Create-time inheritance

When a new file is created, it can inherit an acl from its parent
directory; this is similar to how default acls work in POSIX (draft)
ACLs.

As with POSIX ACLs, if a file inherits an acl from its parent directory,
the intersection between the create mode and the permissions granted by
the inherited acl determines the file masks and file permission bits,
and the umask is ignored.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_base.c | 68 +++++++++++++++++++++++++++++++++++++++++++++++++
fs/richacl_inode.c | 65 ++++++++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 2 ++
3 files changed, 135 insertions(+)

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index 106e988..c1490c7 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -483,3 +483,71 @@ richacl_equiv_mode(const struct richacl *acl, mode_t *mode_p)
return 0;
}
EXPORT_SYMBOL_GPL(richacl_equiv_mode);
+
+/**
+ * richacl_inherit - compute the inherited acl of a new file
+ * @dir_acl: acl of the containing directory
+ * @isdir: inherit by a directory or non-directory?
+ *
+ * A directory can have acl entries which files and/or directories created
+ * inside the directory will inherit. This function computes the acl for such
+ * a new file. If there is no inheritable acl, it will return %NULL.
+ */
+struct richacl *
+richacl_inherit(const struct richacl *dir_acl, int isdir)
+{
+ const struct richace *dir_ace;
+ struct richacl *acl = NULL;
+ struct richace *ace;
+ int count = 0;
+
+ if (isdir) {
+ richacl_for_each_entry(dir_ace, dir_acl) {
+ if (!richace_is_inheritable(dir_ace))
+ continue;
+ count++;
+ }
+ if (!count)
+ return NULL;
+ acl = richacl_alloc(count, GFP_KERNEL);
+ if (!acl)
+ return ERR_PTR(-ENOMEM);
+ ace = acl->a_entries;
+ richacl_for_each_entry(dir_ace, dir_acl) {
+ if (!richace_is_inheritable(dir_ace))
+ continue;
+ richace_copy(ace, dir_ace);
+ if (dir_ace->e_flags & RICHACE_NO_PROPAGATE_INHERIT_ACE)
+ ace->e_flags &= ~RICHACE_INHERITANCE_FLAGS;
+ else if (!(dir_ace->e_flags & RICHACE_DIRECTORY_INHERIT_ACE))
+ ace->e_flags |= RICHACE_INHERIT_ONLY_ACE;
+ ace++;
+ }
+ } else {
+ richacl_for_each_entry(dir_ace, dir_acl) {
+ if (!(dir_ace->e_flags & RICHACE_FILE_INHERIT_ACE))
+ continue;
+ count++;
+ }
+ if (!count)
+ return NULL;
+ acl = richacl_alloc(count, GFP_KERNEL);
+ if (!acl)
+ return ERR_PTR(-ENOMEM);
+ ace = acl->a_entries;
+ richacl_for_each_entry(dir_ace, dir_acl) {
+ if (!(dir_ace->e_flags & RICHACE_FILE_INHERIT_ACE))
+ continue;
+ richace_copy(ace, dir_ace);
+ ace->e_flags &= ~RICHACE_INHERITANCE_FLAGS;
+ /*
+ * RICHACE_DELETE_CHILD is meaningless for
+ * non-directories, so clear it.
+ */
+ ace->e_mask &= ~RICHACE_DELETE_CHILD;
+ ace++;
+ }
+ }
+
+ return acl;
+}
diff --git a/fs/richacl_inode.c b/fs/richacl_inode.c
index dcf8fb1..737f054 100644
--- a/fs/richacl_inode.c
+++ b/fs/richacl_inode.c
@@ -221,3 +221,68 @@ out:
return denied ? -EACCES : 0;
}
EXPORT_SYMBOL_GPL(richacl_permission);
+
+/**
+ * richacl_inherit_inode - compute inherited acl and file mode
+ * @dir_acl: acl of the containing directory
+ * @inode: inode of the new file (create mode in i_mode)
+ *
+ * The file permission bits in inode->i_mode must be set to the create mode by
+ * the caller.
+ *
+ * If there is an inheritable acl, the maximum permissions that the acl grants
+ * will be computed and permissions not granted by the acl will be removed from
+ * inode->i_mode. If there is no inheritable acl, the umask will be applied
+ * instead.
+ */
+static struct richacl *
+richacl_inherit_inode(const struct richacl *dir_acl, struct inode *inode)
+{
+ struct richacl *acl;
+ mode_t mask;
+
+ acl = richacl_inherit(dir_acl, S_ISDIR(inode->i_mode));
+ if (acl) {
+ mask = inode->i_mode;
+ if (richacl_equiv_mode(acl, &mask) == 0) {
+ richacl_put(acl);
+ acl = NULL;
+ } else {
+ richacl_compute_max_masks(acl);
+ /*
+ * Ensure that the acl will not grant any permissions
+ * beyond the create mode.
+ */
+ acl->a_flags |= RICHACL_MASKED;
+ acl->a_owner_mask &=
+ richacl_mode_to_mask(inode->i_mode >> 6);
+ acl->a_group_mask &=
+ richacl_mode_to_mask(inode->i_mode >> 3);
+ acl->a_other_mask &=
+ richacl_mode_to_mask(inode->i_mode);
+ mask = ~S_IRWXUGO | richacl_masks_to_mode(acl);
+ }
+ } else
+ mask = ~current_umask();
+
+ inode->i_mode &= mask;
+ return acl;
+}
+
+struct richacl *richacl_create(struct inode *inode, struct inode *dir)
+{
+ struct richacl *dir_acl, *acl = NULL;
+
+ if (S_ISLNK(inode->i_mode))
+ return NULL;
+ dir_acl = get_richacl(dir);
+ if (dir_acl) {
+ if (IS_ERR(dir_acl))
+ return dir_acl;
+ acl = richacl_inherit_inode(dir_acl, inode);
+ richacl_put(dir_acl);
+ } else
+ inode->i_mode &= ~current_umask();
+ return acl;
+}
+EXPORT_SYMBOL_GPL(richacl_create);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index 6535ce5..9bf95c2 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -305,8 +305,10 @@ extern unsigned int richacl_want_to_mask(unsigned int);
extern void richacl_compute_max_masks(struct richacl *);
extern struct richacl *richacl_chmod(struct richacl *, mode_t);
extern int richacl_equiv_mode(const struct richacl *, mode_t *);
+extern struct richacl *richacl_inherit(const struct richacl *, int);

/* richacl_inode.c */
extern int richacl_permission(struct inode *, const struct richacl *, int);
+extern struct richacl *richacl_create(struct inode *, struct inode *);

#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:11:00

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 13/41] richacl: Check if an acl is equivalent to a file mode

ACLs are considered equivalent to file modes if they only consist of
owner@, group@, and everyone@ entries, the owner@ permissions do not
depend on whether the owner is a member in the owning group, and no
inheritance flags are set. This test is used to avoid storing richacls
if the acl can be computed from the file permission bits.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_base.c | 104 ++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 1 +
2 files changed, 105 insertions(+)

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index 3163152..106e988 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -379,3 +379,107 @@ richacl_chmod(struct richacl *acl, mode_t mode)
return clone;
}
EXPORT_SYMBOL_GPL(richacl_chmod);
+
+/**
+ * richacl_equiv_mode - compute the mode equivalent of @acl
+ *
+ * An acl is considered equivalent to a file mode if it only consists of
+ * owner@, group@, and everyone@ entries and the owner@ permissions do not
+ * depend on whether the owner is a member in the owning group.
+ */
+int
+richacl_equiv_mode(const struct richacl *acl, mode_t *mode_p)
+{
+ mode_t mode = *mode_p;
+
+ /*
+ * The RICHACE_DELETE_CHILD flag is meaningless for non-directories, so
+ * we ignore it.
+ */
+ unsigned int x = S_ISDIR(mode) ? 0 : RICHACE_DELETE_CHILD;
+ struct {
+ unsigned int allowed;
+ unsigned int defined; /* allowed or denied */
+ } owner = {
+ .defined = RICHACE_POSIX_ALWAYS_ALLOWED |
+ RICHACE_POSIX_OWNER_ALLOWED | x,
+ }, group = {
+ .defined = RICHACE_POSIX_ALWAYS_ALLOWED | x,
+ }, everyone = {
+ .defined = RICHACE_POSIX_ALWAYS_ALLOWED | x,
+ };
+ const struct richace *ace;
+
+ if (acl->a_flags & ~(RICHACL_WRITE_THROUGH | RICHACL_MASKED))
+ return -1;
+
+ richacl_for_each_entry(ace, acl) {
+ if (ace->e_flags & ~RICHACE_SPECIAL_WHO)
+ return -1;
+
+ if (richace_is_owner(ace) || richace_is_everyone(ace)) {
+ x = ace->e_mask & ~owner.defined;
+ if (richace_is_allow(ace)) {
+ unsigned int group_denied =
+ group.defined & ~group.allowed;
+
+ if (x & group_denied)
+ return -1;
+ owner.allowed |= x;
+ } else /* if (richace_is_deny(ace)) */ {
+ if (x & group.allowed)
+ return -1;
+ }
+ owner.defined |= x;
+
+ if (richace_is_everyone(ace)) {
+ x = ace->e_mask;
+ if (richace_is_allow(ace)) {
+ group.allowed |=
+ x & ~group.defined;
+ everyone.allowed |=
+ x & ~everyone.defined;
+ }
+ group.defined |= x;
+ everyone.defined |= x;
+ }
+ } else if (richace_is_group(ace)) {
+ x = ace->e_mask & ~group.defined;
+ if (richace_is_allow(ace))
+ group.allowed |= x;
+ group.defined |= x;
+ } else
+ return -1;
+ }
+
+ if (group.allowed & ~owner.defined)
+ return -1;
+
+ if (acl->a_flags & RICHACL_MASKED) {
+ if (acl->a_flags & RICHACL_WRITE_THROUGH) {
+ owner.allowed = acl->a_owner_mask;
+ everyone.allowed = acl->a_other_mask;
+ } else {
+ owner.allowed &= acl->a_owner_mask;
+ everyone.allowed &= acl->a_other_mask;
+ }
+ group.allowed &= acl->a_group_mask;
+ }
+
+ mode = (mode & ~S_IRWXUGO) |
+ (richacl_mask_to_mode(owner.allowed) << 6) |
+ (richacl_mask_to_mode(group.allowed) << 3) |
+ richacl_mask_to_mode(everyone.allowed);
+
+ /* Mask flags we can ignore */
+ x = S_ISDIR(mode) ? 0 : RICHACE_DELETE_CHILD;
+
+ if (((richacl_mode_to_mask(mode >> 6) ^ owner.allowed) & ~x) ||
+ ((richacl_mode_to_mask(mode >> 3) ^ group.allowed) & ~x) ||
+ ((richacl_mode_to_mask(mode) ^ everyone.allowed) & ~x))
+ return -1;
+
+ *mode_p = mode;
+ return 0;
+}
+EXPORT_SYMBOL_GPL(richacl_equiv_mode);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index d4a576c..6535ce5 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -304,6 +304,7 @@ extern unsigned int richacl_mode_to_mask(mode_t);
extern unsigned int richacl_want_to_mask(unsigned int);
extern void richacl_compute_max_masks(struct richacl *);
extern struct richacl *richacl_chmod(struct richacl *, mode_t);
+extern int richacl_equiv_mode(const struct richacl *, mode_t *);

/* richacl_inode.c */
extern int richacl_permission(struct inode *, const struct richacl *, int);
--
2.4.3


2015-09-27 22:10:54

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 12/41] vfs: Cache richacl in struct inode

Cache richacls in struct inode so that this doesn't have to be done
individually in each filesystem. This is similar to POSIX ACLs.

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/inode.c | 11 ++++++--
fs/posix_acl.c | 2 +-
fs/richacl_base.c | 4 +--
fs/richacl_inode.c | 75 +++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/fs.h | 6 +++-
include/linux/richacl.h | 15 ++++++----
6 files changed, 101 insertions(+), 12 deletions(-)

diff --git a/fs/inode.c b/fs/inode.c
index 2a387f4..8462ddb 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -174,8 +174,11 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
inode->i_private = NULL;
inode->i_mapping = mapping;
INIT_HLIST_HEAD(&inode->i_dentry); /* buggered by rcu freeing */
-#ifdef CONFIG_FS_POSIX_ACL
- inode->i_acl = inode->i_default_acl = ACL_NOT_CACHED;
+#if defined(CONFIG_FS_POSIX_ACL) || defined(CONFIG_FS_RICHACL)
+ inode->i_acl = ACL_NOT_CACHED;
+# if defined(CONFIG_FS_POSIX_ACL)
+ inode->i_default_acl = ACL_NOT_CACHED;
+# endif
#endif

#ifdef CONFIG_FSNOTIFY
@@ -231,11 +234,13 @@ void __destroy_inode(struct inode *inode)
atomic_long_dec(&inode->i_sb->s_remove_count);
}

-#ifdef CONFIG_FS_POSIX_ACL
+#if defined(CONFIG_FS_POSIX_ACL) || defined(CONFIG_FS_RICHACL)
if (inode->i_acl && inode->i_acl != ACL_NOT_CACHED)
put_base_acl(inode->i_acl);
+# if defined(CONFIG_FS_POSIX_ACL)
if (inode->i_default_acl && inode->i_default_acl != ACL_NOT_CACHED)
put_base_acl(inode->i_default_acl);
+# endif
#endif
this_cpu_dec(nr_inodes);
}
diff --git a/fs/posix_acl.c b/fs/posix_acl.c
index b3b2265..1d766a5 100644
--- a/fs/posix_acl.c
+++ b/fs/posix_acl.c
@@ -38,7 +38,7 @@ struct posix_acl *get_cached_acl(struct inode *inode, int type)
{
struct posix_acl **p = acl_by_type(inode, type);
struct posix_acl *acl = ACCESS_ONCE(*p);
- if (acl) {
+ if (acl && IS_POSIXACL(inode)) {
spin_lock(&inode->i_lock);
acl = *p;
if (acl != ACL_NOT_CACHED)
diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index 6c69234..3163152 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -33,7 +33,7 @@ richacl_alloc(int count, gfp_t gfp)
struct richacl *acl = kzalloc(size, gfp);

if (acl) {
- atomic_set(&acl->a_refcount, 1);
+ atomic_set(&acl->a_base.ba_refcount, 1);
acl->a_count = count;
}
return acl;
@@ -52,7 +52,7 @@ richacl_clone(const struct richacl *acl, gfp_t gfp)

if (dup) {
memcpy(dup, acl, size);
- atomic_set(&dup->a_refcount, 1);
+ atomic_set(&dup->a_base.ba_refcount, 1);
}
return dup;
}
diff --git a/fs/richacl_inode.c b/fs/richacl_inode.c
index 54e899d..dcf8fb1 100644
--- a/fs/richacl_inode.c
+++ b/fs/richacl_inode.c
@@ -20,6 +20,81 @@
#include <linux/slab.h>
#include <linux/richacl.h>

+struct richacl *get_cached_richacl(struct inode *inode)
+{
+ struct richacl *acl;
+
+ acl = (struct richacl *)ACCESS_ONCE(inode->i_acl);
+ if (acl && IS_RICHACL(inode)) {
+ spin_lock(&inode->i_lock);
+ acl = (struct richacl *)inode->i_acl;
+ if (acl != ACL_NOT_CACHED)
+ acl = richacl_get(acl);
+ spin_unlock(&inode->i_lock);
+ }
+ return acl;
+}
+EXPORT_SYMBOL_GPL(get_cached_richacl);
+
+struct richacl *get_cached_richacl_rcu(struct inode *inode)
+{
+ return (struct richacl *)rcu_dereference(inode->i_acl);
+}
+EXPORT_SYMBOL_GPL(get_cached_richacl_rcu);
+
+void set_cached_richacl(struct inode *inode, struct richacl *acl)
+{
+ struct base_acl *old = NULL;
+
+ spin_lock(&inode->i_lock);
+ old = inode->i_acl;
+ rcu_assign_pointer(inode->i_acl, &richacl_get(acl)->a_base);
+ spin_unlock(&inode->i_lock);
+ if (old != ACL_NOT_CACHED)
+ put_base_acl(old);
+}
+EXPORT_SYMBOL_GPL(set_cached_richacl);
+
+void forget_cached_richacl(struct inode *inode)
+{
+ struct base_acl *old = NULL;
+
+ spin_lock(&inode->i_lock);
+ old = inode->i_acl;
+ inode->i_acl = ACL_NOT_CACHED;
+ spin_unlock(&inode->i_lock);
+ if (old != ACL_NOT_CACHED)
+ put_base_acl(old);
+}
+EXPORT_SYMBOL_GPL(forget_cached_richacl);
+
+struct richacl *get_richacl(struct inode *inode)
+{
+ struct richacl *acl;
+
+ acl = get_cached_richacl(inode);
+ if (acl != ACL_NOT_CACHED)
+ return acl;
+
+ if (!IS_RICHACL(inode))
+ return NULL;
+
+ /*
+ * A filesystem can force a ACL callback by just never filling the
+ * ACL cache. But normally you'd fill the cache either at inode
+ * instantiation time, or on the first ->get_richacl call.
+ *
+ * If the filesystem doesn't have a get_richacl() function at all,
+ * we'll just create the negative cache entry.
+ */
+ if (!inode->i_op->get_richacl) {
+ set_cached_richacl(inode, NULL);
+ return NULL;
+ }
+ return inode->i_op->get_richacl(inode);
+}
+EXPORT_SYMBOL_GPL(get_richacl);
+
/**
* richacl_permission - richacl permission check algorithm
* @inode: inode to check
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 3c22c92..45ff805 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -583,6 +583,7 @@ struct base_acl {
};
};
struct posix_acl;
+struct richacl;
#define ACL_NOT_CACHED ((void *)(-1))

#define IOP_FASTPERM 0x0001
@@ -601,9 +602,11 @@ struct inode {
kgid_t i_gid;
unsigned int i_flags;

-#if defined(CONFIG_FS_POSIX_ACL)
+#if defined(CONFIG_FS_POSIX_ACL) || defined(CONFIG_FS_RICHACL)
struct base_acl *i_acl;
+# if defined(CONFIG_FS_POSIX_ACL)
struct base_acl *i_default_acl;
+# endif
#endif

const struct inode_operations *i_op;
@@ -1662,6 +1665,7 @@ struct inode_operations {
const char * (*follow_link) (struct dentry *, void **);
int (*permission) (struct inode *, int);
struct posix_acl * (*get_acl)(struct inode *, int);
+ struct richacl * (*get_richacl)(struct inode *);

int (*readlink) (struct dentry *, char __user *,int);
void (*put_link) (struct inode *, void *);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index 61836f1..d4a576c 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -33,7 +33,7 @@ struct richace {
};

struct richacl {
- atomic_t a_refcount;
+ struct base_acl a_base;
unsigned int a_owner_mask;
unsigned int a_group_mask;
unsigned int a_other_mask;
@@ -175,8 +175,7 @@ struct richacl {
static inline struct richacl *
richacl_get(struct richacl *acl)
{
- if (acl)
- atomic_inc(&acl->a_base.ba_refcount);
+ get_base_acl(&acl->a_base);
return acl;
}

@@ -186,10 +185,16 @@ richacl_get(struct richacl *acl)
static inline void
richacl_put(struct richacl *acl)
{
- if (acl && atomic_dec_and_test(&acl->a_refcount))
- kfree(acl);
+ BUILD_BUG_ON(offsetof(struct richacl, a_base) != 0);
+ put_base_acl(&acl->a_base);
}

+extern struct richacl *get_cached_richacl(struct inode *);
+extern struct richacl *get_cached_richacl_rcu(struct inode *);
+extern void set_cached_richacl(struct inode *, struct richacl *);
+extern void forget_cached_richacl(struct inode *);
+extern struct richacl *get_richacl(struct inode *);
+
/**
* richace_is_owner - check if @ace is an OWNER@ entry
*/
--
2.4.3


2015-09-27 22:10:48

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 11/41] vfs: Cache base_acl objects in inodes

POSIX ACLs and richacls are both objects allocated by kmalloc() with a
reference count which are freed by kfree_rcu(). An inode can either
cache an access and a default POSIX ACL, or a richacl (richacls do not
have default acls). To allow an inode to cache either of the two kinds
of acls, introduce a new base_acl type and convert i_acl and
i_default_acl to that type. In most cases, the vfs then doesn't have to
care which kind of acl an inode caches (if any).

Signed-off-by: Andreas Gruenbacher <[email protected]>
---
drivers/staging/lustre/lustre/llite/llite_lib.c | 2 +-
fs/f2fs/acl.c | 4 ++--
fs/inode.c | 4 ++--
fs/jffs2/acl.c | 6 ++++--
fs/posix_acl.c | 18 +++++++++---------
include/linux/fs.h | 25 ++++++++++++++++++++++---
include/linux/posix_acl.h | 12 ++++--------
include/linux/richacl.h | 2 +-
8 files changed, 45 insertions(+), 28 deletions(-)

diff --git a/drivers/staging/lustre/lustre/llite/llite_lib.c b/drivers/staging/lustre/lustre/llite/llite_lib.c
index b4ed6c8..5766f69 100644
--- a/drivers/staging/lustre/lustre/llite/llite_lib.c
+++ b/drivers/staging/lustre/lustre/llite/llite_lib.c
@@ -1118,7 +1118,7 @@ void ll_clear_inode(struct inode *inode)
}
#ifdef CONFIG_FS_POSIX_ACL
else if (lli->lli_posix_acl) {
- LASSERT(atomic_read(&lli->lli_posix_acl->a_refcount) == 1);
+ LASSERT(atomic_read(&lli->lli_posix_acl->a_base.ba_refcount) == 1);
LASSERT(lli->lli_remote_perms == NULL);
posix_acl_release(lli->lli_posix_acl);
lli->lli_posix_acl = NULL;
diff --git a/fs/f2fs/acl.c b/fs/f2fs/acl.c
index c8f25f7..a4207de 100644
--- a/fs/f2fs/acl.c
+++ b/fs/f2fs/acl.c
@@ -270,7 +270,7 @@ static struct posix_acl *f2fs_acl_clone(const struct posix_acl *acl,
sizeof(struct posix_acl_entry);
clone = kmemdup(acl, size, flags);
if (clone)
- atomic_set(&clone->a_refcount, 1);
+ atomic_set(&clone->a_base.ba_refcount, 1);
}
return clone;
}
@@ -282,7 +282,7 @@ static int f2fs_acl_create_masq(struct posix_acl *acl, umode_t *mode_p)
umode_t mode = *mode_p;
int not_equiv = 0;

- /* assert(atomic_read(acl->a_refcount) == 1); */
+ /* assert(atomic_read(acl->a_base.ba_refcount) == 1); */

FOREACH_ACL_ENTRY(pa, acl, pe) {
switch(pa->e_tag) {
diff --git a/fs/inode.c b/fs/inode.c
index 78a17b8..2a387f4 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -233,9 +233,9 @@ void __destroy_inode(struct inode *inode)

#ifdef CONFIG_FS_POSIX_ACL
if (inode->i_acl && inode->i_acl != ACL_NOT_CACHED)
- posix_acl_release(inode->i_acl);
+ put_base_acl(inode->i_acl);
if (inode->i_default_acl && inode->i_default_acl != ACL_NOT_CACHED)
- posix_acl_release(inode->i_default_acl);
+ put_base_acl(inode->i_default_acl);
#endif
this_cpu_dec(nr_inodes);
}
diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c
index 2f7a3c0..04a5836 100644
--- a/fs/jffs2/acl.c
+++ b/fs/jffs2/acl.c
@@ -294,13 +294,15 @@ int jffs2_init_acl_post(struct inode *inode)
int rc;

if (inode->i_default_acl) {
- rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_DEFAULT, inode->i_default_acl);
+ rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_DEFAULT,
+ *acl_by_type(inode, ACL_TYPE_DEFAULT));
if (rc)
return rc;
}

if (inode->i_acl) {
- rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_ACCESS, inode->i_acl);
+ rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_ACCESS,
+ *acl_by_type(inode, ACL_TYPE_ACCESS));
if (rc)
return rc;
}
diff --git a/fs/posix_acl.c b/fs/posix_acl.c
index 4fb17de..b3b2265 100644
--- a/fs/posix_acl.c
+++ b/fs/posix_acl.c
@@ -25,9 +25,9 @@ struct posix_acl **acl_by_type(struct inode *inode, int type)
{
switch (type) {
case ACL_TYPE_ACCESS:
- return &inode->i_acl;
+ return (struct posix_acl **)&inode->i_acl;
case ACL_TYPE_DEFAULT:
- return &inode->i_default_acl;
+ return (struct posix_acl **)&inode->i_default_acl;
default:
BUG();
}
@@ -83,16 +83,16 @@ EXPORT_SYMBOL(forget_cached_acl);

void forget_all_cached_acls(struct inode *inode)
{
- struct posix_acl *old_access, *old_default;
+ struct base_acl *old_access, *old_default;
spin_lock(&inode->i_lock);
old_access = inode->i_acl;
old_default = inode->i_default_acl;
inode->i_acl = inode->i_default_acl = ACL_NOT_CACHED;
spin_unlock(&inode->i_lock);
if (old_access != ACL_NOT_CACHED)
- posix_acl_release(old_access);
+ put_base_acl(old_access);
if (old_default != ACL_NOT_CACHED)
- posix_acl_release(old_default);
+ put_base_acl(old_default);
}
EXPORT_SYMBOL(forget_all_cached_acls);

@@ -129,7 +129,7 @@ EXPORT_SYMBOL(get_acl);
void
posix_acl_init(struct posix_acl *acl, int count)
{
- atomic_set(&acl->a_refcount, 1);
+ atomic_set(&acl->a_base.ba_refcount, 1);
acl->a_count = count;
}
EXPORT_SYMBOL(posix_acl_init);
@@ -162,7 +162,7 @@ posix_acl_clone(const struct posix_acl *acl, gfp_t flags)
sizeof(struct posix_acl_entry);
clone = kmemdup(acl, size, flags);
if (clone)
- atomic_set(&clone->a_refcount, 1);
+ atomic_set(&clone->a_base.ba_refcount, 1);
}
return clone;
}
@@ -384,7 +384,7 @@ static int posix_acl_create_masq(struct posix_acl *acl, umode_t *mode_p)
umode_t mode = *mode_p;
int not_equiv = 0;

- /* assert(atomic_read(acl->a_refcount) == 1); */
+ /* assert(atomic_read(acl->a_base.ba_refcount) == 1); */

FOREACH_ACL_ENTRY(pa, acl, pe) {
switch(pa->e_tag) {
@@ -439,7 +439,7 @@ static int __posix_acl_chmod_masq(struct posix_acl *acl, umode_t mode)
struct posix_acl_entry *group_obj = NULL, *mask_obj = NULL;
struct posix_acl_entry *pa, *pe;

- /* assert(atomic_read(acl->a_refcount) == 1); */
+ /* assert(atomic_read(acl->a_base.ba_refcount) == 1); */

FOREACH_ACL_ENTRY(pa, acl, pe) {
switch(pa->e_tag) {
diff --git a/include/linux/fs.h b/include/linux/fs.h
index ba91a89..3c22c92 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -576,6 +576,12 @@ static inline void mapping_allow_writable(struct address_space *mapping)
#define i_size_ordered_init(inode) do { } while (0)
#endif

+struct base_acl {
+ union {
+ atomic_t ba_refcount;
+ struct rcu_head ba_rcu;
+ };
+};
struct posix_acl;
#define ACL_NOT_CACHED ((void *)(-1))

@@ -595,9 +601,9 @@ struct inode {
kgid_t i_gid;
unsigned int i_flags;

-#ifdef CONFIG_FS_POSIX_ACL
- struct posix_acl *i_acl;
- struct posix_acl *i_default_acl;
+#if defined(CONFIG_FS_POSIX_ACL)
+ struct base_acl *i_acl;
+ struct base_acl *i_default_acl;
#endif

const struct inode_operations *i_op;
@@ -3059,4 +3065,17 @@ static inline bool dir_relax(struct inode *inode)

extern bool path_noexec(const struct path *path);

+static inline struct base_acl *get_base_acl(struct base_acl *acl)
+{
+ if (acl)
+ atomic_inc(&acl->ba_refcount);
+ return acl;
+}
+
+static inline void put_base_acl(struct base_acl *acl)
+{
+ if (acl && atomic_dec_and_test(&acl->ba_refcount))
+ kfree_rcu(acl, ba_rcu);
+}
+
#endif /* _LINUX_FS_H */
diff --git a/include/linux/posix_acl.h b/include/linux/posix_acl.h
index 3e96a6a..2c46441 100644
--- a/include/linux/posix_acl.h
+++ b/include/linux/posix_acl.h
@@ -43,10 +43,7 @@ struct posix_acl_entry {
};

struct posix_acl {
- union {
- atomic_t a_refcount;
- struct rcu_head a_rcu;
- };
+ struct base_acl a_base;
unsigned int a_count;
struct posix_acl_entry a_entries[0];
};
@@ -61,8 +58,7 @@ struct posix_acl {
static inline struct posix_acl *
posix_acl_dup(struct posix_acl *acl)
{
- if (acl)
- atomic_inc(&acl->a_refcount);
+ get_base_acl(&acl->a_base);
return acl;
}

@@ -72,8 +68,8 @@ posix_acl_dup(struct posix_acl *acl)
static inline void
posix_acl_release(struct posix_acl *acl)
{
- if (acl && atomic_dec_and_test(&acl->a_refcount))
- kfree_rcu(acl, a_rcu);
+ BUILD_BUG_ON(offsetof(struct posix_acl, a_base) != 0);
+ put_base_acl(&acl->a_base);
}


diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index 9768eeb..61836f1 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -176,7 +176,7 @@ static inline struct richacl *
richacl_get(struct richacl *acl)
{
if (acl)
- atomic_inc(&acl->a_refcount);
+ atomic_inc(&acl->a_base.ba_refcount);
return acl;
}

--
2.4.3


2015-09-27 22:10:43

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 10/41] richacl: Permission check algorithm

A richacl roughly grants a requested access if the NFSv4 acl in the
richacl grants the requested permissions according to the NFSv4
permission check algorithm and the file mask that applies to the process
includes the requested permissions.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: "J. Bruce Fields" <[email protected]>
---
fs/Makefile | 2 +-
fs/richacl_inode.c | 148 ++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 3 +
3 files changed, 152 insertions(+), 1 deletion(-)
create mode 100644 fs/richacl_inode.c

diff --git a/fs/Makefile b/fs/Makefile
index fe3e9dd..ec665fd 100644
--- a/fs/Makefile
+++ b/fs/Makefile
@@ -49,7 +49,7 @@ obj-$(CONFIG_SYSCTL) += drop_caches.o

obj-$(CONFIG_FHANDLE) += fhandle.o
obj-$(CONFIG_FS_RICHACL) += richacl.o
-richacl-y := richacl_base.o
+richacl-y := richacl_base.o richacl_inode.o

obj-y += quota/

diff --git a/fs/richacl_inode.c b/fs/richacl_inode.c
new file mode 100644
index 0000000..54e899d
--- /dev/null
+++ b/fs/richacl_inode.c
@@ -0,0 +1,148 @@
+/*
+ * Copyright (C) 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/sched.h>
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/richacl.h>
+
+/**
+ * richacl_permission - richacl permission check algorithm
+ * @inode: inode to check
+ * @acl: rich acl of the inode
+ * @want: requested access (MAY_* flags)
+ *
+ * Checks if the current process is granted @mask flags in @acl.
+ */
+int
+richacl_permission(struct inode *inode, const struct richacl *acl,
+ int want)
+{
+ const struct richace *ace;
+ unsigned int mask = richacl_want_to_mask(want);
+ unsigned int requested = mask, denied = 0;
+ int in_owning_group = in_group_p(inode->i_gid);
+ int in_owner_or_group_class = in_owning_group;
+
+ /*
+ * A process is
+ * - in the owner file class if it owns the file,
+ * - in the group file class if it is in the file's owning group or
+ * it matches any of the user or group entries, and
+ * - in the other file class otherwise.
+ * The file class is only relevant for determining which file mask to
+ * apply, which only happens for masked acls.
+ */
+ if (acl->a_flags & RICHACL_MASKED) {
+ if ((acl->a_flags & RICHACL_WRITE_THROUGH) &&
+ uid_eq(current_fsuid(), inode->i_uid)) {
+ denied = requested & ~acl->a_owner_mask;
+ goto out;
+ }
+ } else {
+ /*
+ * When the acl is not masked, there is no need to determine if
+ * the process is in the group class and we can break out
+ * earlier of the loop below.
+ */
+ in_owner_or_group_class = 1;
+ }
+
+ /*
+ * Check if the acl grants the requested access and determine which
+ * file class the process is in.
+ */
+ richacl_for_each_entry(ace, acl) {
+ unsigned int ace_mask = ace->e_mask;
+
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_owner(ace)) {
+ if (!uid_eq(current_fsuid(), inode->i_uid))
+ continue;
+ goto entry_matches_owner;
+ } else if (richace_is_group(ace)) {
+ if (!in_owning_group)
+ continue;
+ } else if (richace_is_unix_user(ace)) {
+ if (!uid_eq(current_fsuid(), ace->e_id.uid))
+ continue;
+ goto entry_matches_owner;
+ } else if (richace_is_unix_group(ace)) {
+ if (!in_group_p(ace->e_id.gid))
+ continue;
+ } else
+ goto entry_matches_everyone;
+
+ /*
+ * Apply the group file mask to entries other than owner@ and
+ * everyone@ or user entries matching the owner. This ensures
+ * that we grant the same permissions as the acl computed by
+ * richacl_apply_masks().
+ *
+ * Without this restriction, the following richacl would grant
+ * rw access to processes which are both the owner and in the
+ * owning group, but not to other users in the owning group,
+ * which could not be represented without masks:
+ *
+ * owner:rw::mask
+ * group@:rw::allow
+ */
+ if ((acl->a_flags & RICHACL_MASKED) && richace_is_allow(ace))
+ ace_mask &= acl->a_group_mask;
+
+entry_matches_owner:
+ /* The process is in the owner or group file class. */
+ in_owner_or_group_class = 1;
+
+entry_matches_everyone:
+ /* Check which mask flags the ACE allows or denies. */
+ if (richace_is_deny(ace))
+ denied |= ace_mask & mask;
+ mask &= ~ace_mask;
+
+ /*
+ * Keep going until we know which file class
+ * the process is in.
+ */
+ if (!mask && in_owner_or_group_class)
+ break;
+ }
+ denied |= mask;
+
+ if (acl->a_flags & RICHACL_MASKED) {
+ /*
+ * The file class a process is in determines which file mask
+ * applies. Check if that file mask also grants the requested
+ * access.
+ */
+ if (uid_eq(current_fsuid(), inode->i_uid))
+ denied |= requested & ~acl->a_owner_mask;
+ else if (in_owner_or_group_class)
+ denied |= requested & ~acl->a_group_mask;
+ else {
+ if (acl->a_flags & RICHACL_WRITE_THROUGH)
+ denied = requested & ~acl->a_other_mask;
+ else
+ denied |= requested & ~acl->a_other_mask;
+ }
+ }
+
+out:
+ return denied ? -EACCES : 0;
+}
+EXPORT_SYMBOL_GPL(richacl_permission);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index e00f313..9768eeb 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -300,4 +300,7 @@ extern unsigned int richacl_want_to_mask(unsigned int);
extern void richacl_compute_max_masks(struct richacl *);
extern struct richacl *richacl_chmod(struct richacl *, mode_t);

+/* richacl_inode.c */
+extern int richacl_permission(struct inode *, const struct richacl *, int);
+
#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:10:37

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 09/41] richacl: Update the file masks in chmod()

Doing a chmod() sets the file mode, which includes the file permission
bits. When a file has a richacl, the permissions that the richacl
grants need to be limited to what the new file permission bits allow.

This is done by setting the file masks in the richacl to what the file
permission bits map to. The richacl access check algorithm takes the
file masks into account, which ensures that the richacl cannot grant too
many permissions.

It is possible to explicitly add permissions to the file masks which go
beyond what the file permission bits can grant (like the
RICHACE_WRITE_ACL permission). The POSIX.1 standard calls this an
alternate file access control mechanism. A subsequent chmod() would
ensure that those permissions are disabled again.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_base.c | 40 ++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 1 +
2 files changed, 41 insertions(+)

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index fc544f7..6c69234 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -339,3 +339,43 @@ restart:
acl->a_flags &= ~(RICHACL_WRITE_THROUGH | RICHACL_MASKED);
}
EXPORT_SYMBOL_GPL(richacl_compute_max_masks);
+
+/**
+ * richacl_chmod - update the file masks to reflect the new mode
+ * @mode: new file permission bits including the file type
+ *
+ * Return a copy of @acl where the file masks have been replaced by the file
+ * masks corresponding to the file permission bits in @mode, or returns @acl
+ * itself if the file masks are already up to date. Takes over a reference
+ * to @acl.
+ */
+struct richacl *
+richacl_chmod(struct richacl *acl, mode_t mode)
+{
+ unsigned int x = S_ISDIR(mode) ? 0 : RICHACE_DELETE_CHILD;
+ unsigned int owner_mask, group_mask, other_mask;
+ struct richacl *clone;
+
+ owner_mask = richacl_mode_to_mask(mode >> 6) & ~x;
+ group_mask = richacl_mode_to_mask(mode >> 3) & ~x;
+ other_mask = richacl_mode_to_mask(mode) & ~x;
+
+ if (acl->a_owner_mask == owner_mask &&
+ acl->a_group_mask == group_mask &&
+ acl->a_other_mask == other_mask &&
+ (acl->a_flags & (RICHACL_WRITE_THROUGH | RICHACL_MASKED)))
+ return acl;
+
+ clone = richacl_clone(acl, GFP_KERNEL);
+ richacl_put(acl);
+ if (!clone)
+ return ERR_PTR(-ENOMEM);
+
+ clone->a_flags |= (RICHACL_WRITE_THROUGH | RICHACL_MASKED);
+ clone->a_owner_mask = owner_mask;
+ clone->a_group_mask = group_mask;
+ clone->a_other_mask = other_mask;
+
+ return clone;
+}
+EXPORT_SYMBOL_GPL(richacl_chmod);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index e81144a..e00f313 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -298,5 +298,6 @@ extern int richacl_masks_to_mode(const struct richacl *);
extern unsigned int richacl_mode_to_mask(mode_t);
extern unsigned int richacl_want_to_mask(unsigned int);
extern void richacl_compute_max_masks(struct richacl *);
+extern struct richacl *richacl_chmod(struct richacl *, mode_t);

#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:10:31

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 08/41] richacl: Compute maximum file masks from an acl

Compute upper bound owner, group, and other file masks with as few
permissions as possible without denying any permissions that the NFSv4
acl in a richacl grants.

This algorithm is used when a file inherits an acl at create time and
when an acl is set via a mechanism that does not provide file masks
(such as setting an acl via nfsd). When user-space sets an acl via
setxattr, the extended attribute already includes the file masks.

Setting an acl also sets the file mode permission bits: they are
determined by the file masks; see richacl_masks_to_mode().

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_base.c | 157 ++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 1 +
2 files changed, 158 insertions(+)

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index 063dbe4..fc544f7 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -182,3 +182,160 @@ richacl_want_to_mask(unsigned int want)
return mask;
}
EXPORT_SYMBOL_GPL(richacl_want_to_mask);
+
+/*
+ * Note: functions like richacl_allowed_to_who(), richacl_group_class_allowed(),
+ * and richacl_compute_max_masks() iterate through the entire acl in reverse
+ * order as an optimization.
+ *
+ * In the standard algorithm, aces are considered in forward order. When a
+ * process matches an ace, the permissions in the ace are either allowed or
+ * denied depending on the ace type. Once a permission has been allowed or
+ * denied, it is no longer considered in further aces.
+ *
+ * By iterating through the acl in reverse order, we can compute the same
+ * result without having to keep track of which permissions have been allowed
+ * and denied already.
+ */
+
+/**
+ * richacl_allowed_to_who - permissions allowed to a specific who value
+ *
+ * Compute the maximum mask values allowed to a specific who value, taking
+ * everyone@ aces into account.
+ */
+static unsigned int richacl_allowed_to_who(struct richacl *acl,
+ struct richace *who)
+{
+ struct richace *ace;
+ unsigned int allowed = 0;
+
+ richacl_for_each_entry_reverse(ace, acl) {
+ if (richace_is_inherit_only(ace))
+ continue;
+ if (richace_is_same_identifier(ace, who) ||
+ richace_is_everyone(ace)) {
+ if (richace_is_allow(ace))
+ allowed |= ace->e_mask;
+ else if (richace_is_deny(ace))
+ allowed &= ~ace->e_mask;
+ }
+ }
+ return allowed;
+}
+
+/**
+ * richacl_group_class_allowed - maximum permissions of the group class
+ *
+ * Compute the maximum mask values allowed to a process in the group class
+ * (i.e., a process which is not the owner but is in the owning group or
+ * matches a user or group acl entry). This includes permissions granted or
+ * denied by everyone@ aces.
+ *
+ * See richacl_compute_max_masks().
+ */
+static unsigned int richacl_group_class_allowed(struct richacl *acl)
+{
+ struct richace *ace;
+ unsigned int everyone_allowed = 0, group_class_allowed = 0;
+ int had_group_ace = 0;
+
+ richacl_for_each_entry_reverse(ace, acl) {
+ if (richace_is_inherit_only(ace) ||
+ richace_is_owner(ace))
+ continue;
+
+ if (richace_is_everyone(ace)) {
+ if (richace_is_allow(ace))
+ everyone_allowed |= ace->e_mask;
+ else if (richace_is_deny(ace))
+ everyone_allowed &= ~ace->e_mask;
+ } else {
+ group_class_allowed |=
+ richacl_allowed_to_who(acl, ace);
+
+ if (richace_is_group(ace))
+ had_group_ace = 1;
+ }
+ }
+ /*
+ * If the acl doesn't contain any group@ aces, richacl_allowed_to_who()
+ * wasn't called for the owning group. We could make that call now, but
+ * we already know the result (everyone_allowed).
+ */
+ if (!had_group_ace)
+ group_class_allowed |= everyone_allowed;
+ return group_class_allowed;
+}
+
+/**
+ * richacl_compute_max_masks - compute upper bound masks
+ *
+ * Computes upper bound owner, group, and other masks so that none of the
+ * permissions allowed by the acl are disabled.
+ *
+ * We don't make assumptions about who the owner is so that the owner can
+ * change with no effect on the file masks or file mode permission bits; this
+ * means that we must assume that all entries can match the owner.
+ */
+void richacl_compute_max_masks(struct richacl *acl)
+{
+ unsigned int gmask = ~0;
+ struct richace *ace;
+
+ /*
+ * @gmask contains all permissions which the group class is ever
+ * allowed. We use it to avoid adding permissions to the group mask
+ * from everyone@ allow aces which the group class is always denied
+ * through other aces. For example, the following acl would otherwise
+ * result in a group mask of rw:
+ *
+ * group@:w::deny
+ * everyone@:rw::allow
+ *
+ * Avoid computing @gmask for acls which do not include any group class
+ * deny aces: in such acls, the group class is never denied any
+ * permissions from everyone@ allow aces, and the group class cannot
+ * have fewer permissions than the other class.
+ */
+
+restart:
+ acl->a_owner_mask = 0;
+ acl->a_group_mask = 0;
+ acl->a_other_mask = 0;
+
+ richacl_for_each_entry_reverse(ace, acl) {
+ if (richace_is_inherit_only(ace))
+ continue;
+
+ if (richace_is_owner(ace)) {
+ if (richace_is_allow(ace))
+ acl->a_owner_mask |= ace->e_mask;
+ else if (richace_is_deny(ace))
+ acl->a_owner_mask &= ~ace->e_mask;
+ } else if (richace_is_everyone(ace)) {
+ if (richace_is_allow(ace)) {
+ acl->a_owner_mask |= ace->e_mask;
+ acl->a_group_mask |= ace->e_mask & gmask;
+ acl->a_other_mask |= ace->e_mask;
+ } else if (richace_is_deny(ace)) {
+ acl->a_owner_mask &= ~ace->e_mask;
+ acl->a_group_mask &= ~ace->e_mask;
+ acl->a_other_mask &= ~ace->e_mask;
+ }
+ } else {
+ if (richace_is_allow(ace)) {
+ acl->a_owner_mask |= ace->e_mask & gmask;
+ acl->a_group_mask |= ace->e_mask & gmask;
+ } else if (richace_is_deny(ace) && gmask == ~0) {
+ gmask = richacl_group_class_allowed(acl);
+ if (likely(gmask != ~0))
+ /* should always be true */
+ goto restart;
+ }
+ }
+ }
+
+ acl->a_flags &= ~(RICHACL_WRITE_THROUGH | RICHACL_MASKED);
+}
+EXPORT_SYMBOL_GPL(richacl_compute_max_masks);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index 9c8f298..e81144a 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -297,5 +297,6 @@ extern void richace_copy(struct richace *, const struct richace *);
extern int richacl_masks_to_mode(const struct richacl *);
extern unsigned int richacl_mode_to_mask(mode_t);
extern unsigned int richacl_want_to_mask(unsigned int);
+extern void richacl_compute_max_masks(struct richacl *);

#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:10:25

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 07/41] richacl: Permission mapping functions

We need to map from POSIX permissions to NFSv4 permissions when a
chmod() is done, from NFSv4 permissions to POSIX permissions when an acl
is set (which implicitly sets the file permission bits), and from the
MAY_READ/MAY_WRITE/MAY_EXEC/MAY_APPEND flags to NFSv4 permissions when
doing an access check in a richacl.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/richacl_base.c | 117 ++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/richacl.h | 47 ++++++++++++++++++-
2 files changed, 163 insertions(+), 1 deletion(-)

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
index 6d9a073..063dbe4 100644
--- a/fs/richacl_base.c
+++ b/fs/richacl_base.c
@@ -65,3 +65,120 @@ richace_copy(struct richace *to, const struct richace *from)
{
memcpy(to, from, sizeof(struct richace));
}
+
+/*
+ * richacl_mask_to_mode - compute the file permission bits from mask
+ * @mask: %RICHACE_* permission mask
+ *
+ * Compute the file permission bits corresponding to a particular set of
+ * richacl permissions.
+ *
+ * See richacl_masks_to_mode().
+ */
+static int
+richacl_mask_to_mode(unsigned int mask)
+{
+ int mode = 0;
+
+ if (mask & RICHACE_POSIX_MODE_READ)
+ mode |= S_IROTH;
+ if (mask & RICHACE_POSIX_MODE_WRITE)
+ mode |= S_IWOTH;
+ if (mask & RICHACE_POSIX_MODE_EXEC)
+ mode |= S_IXOTH;
+
+ return mode;
+}
+
+/**
+ * richacl_masks_to_mode - compute file permission bits from file masks
+ *
+ * When setting a richacl, we set the file permission bits to indicate maximum
+ * permissions: for example, we set the Write permission when a mask contains
+ * RICHACE_APPEND_DATA even if it does not also contain RICHACE_WRITE_DATA.
+ *
+ * Permissions which are not in RICHACE_POSIX_MODE_READ,
+ * RICHACE_POSIX_MODE_WRITE, or RICHACE_POSIX_MODE_EXEC cannot be represented
+ * in the file permission bits. Such permissions can still be effective, but
+ * not for new files or after a chmod(); they must be explicitly enabled in the
+ * richacl.
+ */
+int
+richacl_masks_to_mode(const struct richacl *acl)
+{
+ return richacl_mask_to_mode(acl->a_owner_mask) << 6 |
+ richacl_mask_to_mode(acl->a_group_mask) << 3 |
+ richacl_mask_to_mode(acl->a_other_mask);
+}
+EXPORT_SYMBOL_GPL(richacl_masks_to_mode);
+
+/**
+ * richacl_mode_to_mask - compute a file mask from the lowest three mode bits
+ *
+ * When the file permission bits of a file are set with chmod(), this specifies
+ * the maximum permissions that processes will get. All permissions beyond
+ * that will be removed from the file masks, and become ineffective.
+ */
+unsigned int
+richacl_mode_to_mask(mode_t mode)
+{
+ unsigned int mask = 0;
+
+ if (mode & S_IROTH)
+ mask |= RICHACE_POSIX_MODE_READ;
+ if (mode & S_IWOTH)
+ mask |= RICHACE_POSIX_MODE_WRITE;
+ if (mode & S_IXOTH)
+ mask |= RICHACE_POSIX_MODE_EXEC;
+
+ return mask;
+}
+
+/**
+ * richacl_want_to_mask - convert the iop->permission want argument to a mask
+ * @want: @want argument of the permission inode operation
+ *
+ * When checking for append, @want is (MAY_WRITE | MAY_APPEND).
+ *
+ * Richacls use the iop->may_create and iop->may_delete hooks which are used
+ * for checking if creating and deleting files is allowed. These hooks do not
+ * use richacl_want_to_mask(), so we do not have to deal with mapping MAY_WRITE
+ * to RICHACE_ADD_FILE, RICHACE_ADD_SUBDIRECTORY, and RICHACE_DELETE_CHILD
+ * here.
+ */
+unsigned int
+richacl_want_to_mask(unsigned int want)
+{
+ unsigned int mask = 0;
+
+ if (want & MAY_READ)
+ mask |= RICHACE_READ_DATA;
+ if (want & MAY_DELETE_SELF)
+ mask |= RICHACE_DELETE;
+ if (want & MAY_TAKE_OWNERSHIP)
+ mask |= RICHACE_WRITE_OWNER;
+ if (want & MAY_CHMOD)
+ mask |= RICHACE_WRITE_ACL;
+ if (want & MAY_SET_TIMES)
+ mask |= RICHACE_WRITE_ATTRIBUTES;
+ if (want & MAY_EXEC)
+ mask |= RICHACE_EXECUTE;
+ /*
+ * differentiate MAY_WRITE from these request
+ */
+ if (want & (MAY_APPEND |
+ MAY_CREATE_FILE | MAY_CREATE_DIR |
+ MAY_DELETE_CHILD)) {
+ if (want & MAY_APPEND)
+ mask |= RICHACE_APPEND_DATA;
+ if (want & MAY_CREATE_FILE)
+ mask |= RICHACE_ADD_FILE;
+ if (want & MAY_CREATE_DIR)
+ mask |= RICHACE_ADD_SUBDIRECTORY;
+ if (want & MAY_DELETE_CHILD)
+ mask |= RICHACE_DELETE_CHILD;
+ } else if (want & MAY_WRITE)
+ mask |= RICHACE_WRITE_DATA;
+ return mask;
+}
+EXPORT_SYMBOL_GPL(richacl_want_to_mask);
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
index bfa94bb..9c8f298 100644
--- a/include/linux/richacl.h
+++ b/include/linux/richacl.h
@@ -126,6 +126,49 @@ struct richacl {
RICHACE_WRITE_OWNER | \
RICHACE_SYNCHRONIZE)

+/*
+ * The POSIX permissions are supersets of the following NFSv4 permissions:
+ *
+ * - MAY_READ maps to READ_DATA or LIST_DIRECTORY, depending on the type
+ * of the file system object.
+ *
+ * - MAY_WRITE maps to WRITE_DATA or RICHACE_APPEND_DATA for files, and to
+ * ADD_FILE, RICHACE_ADD_SUBDIRECTORY, or RICHACE_DELETE_CHILD for directories.
+ *
+ * - MAY_EXECUTE maps to RICHACE_EXECUTE.
+ *
+ * (Some of these NFSv4 permissions have the same bit values.)
+ */
+#define RICHACE_POSIX_MODE_READ ( \
+ RICHACE_READ_DATA | \
+ RICHACE_LIST_DIRECTORY)
+#define RICHACE_POSIX_MODE_WRITE ( \
+ RICHACE_WRITE_DATA | \
+ RICHACE_ADD_FILE | \
+ RICHACE_APPEND_DATA | \
+ RICHACE_ADD_SUBDIRECTORY | \
+ RICHACE_DELETE_CHILD)
+#define RICHACE_POSIX_MODE_EXEC RICHACE_EXECUTE
+#define RICHACE_POSIX_MODE_ALL ( \
+ RICHACE_POSIX_MODE_READ | \
+ RICHACE_POSIX_MODE_WRITE | \
+ RICHACE_POSIX_MODE_EXEC)
+/*
+ * These permissions are always allowed
+ * no matter what the acl says.
+ */
+#define RICHACE_POSIX_ALWAYS_ALLOWED ( \
+ RICHACE_SYNCHRONIZE | \
+ RICHACE_READ_ATTRIBUTES | \
+ RICHACE_READ_ACL)
+/*
+ * The owner is implicitly granted
+ * these permissions under POSIX.
+ */
+#define RICHACE_POSIX_OWNER_ALLOWED ( \
+ RICHACE_WRITE_ATTRIBUTES | \
+ RICHACE_WRITE_OWNER | \
+ RICHACE_WRITE_ACL)
/**
* richacl_get - grab another reference to a richacl handle
*/
@@ -251,6 +294,8 @@ richace_is_same_identifier(const struct richace *a, const struct richace *b)
extern struct richacl *richacl_alloc(int, gfp_t);
extern struct richacl *richacl_clone(const struct richacl *, gfp_t);
extern void richace_copy(struct richace *, const struct richace *);
-
+extern int richacl_masks_to_mode(const struct richacl *);
+extern unsigned int richacl_mode_to_mask(mode_t);
+extern unsigned int richacl_want_to_mask(unsigned int);

#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:10:20

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 06/41] richacl: In-memory representation and helper functions

A richacl consists of an NFSv4 acl and an owner, group, and other mask.
These three masks correspond to the owner, group, and other file
permission bits, but they contain NFSv4 permissions instead of POSIX
permissions.

Each entry in the NFSv4 acl applies to the file owner (OWNER@), the
owning group (GROUP@), everyone (EVERYONE@), or to a specific uid or
gid.

As in the standard POSIX file permission model, each process is the
owner, group, or other file class. A richacl grants a requested access
only if the NFSv4 acl in the richacl grants the access (according to the
NFSv4 permission check algorithm), and the file mask that applies to the
process includes the requested permissions.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/Makefile | 2 +
fs/richacl_base.c | 67 +++++++++++++
include/linux/richacl.h | 256 ++++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 325 insertions(+)
create mode 100644 fs/richacl_base.c
create mode 100644 include/linux/richacl.h

diff --git a/fs/Makefile b/fs/Makefile
index f79cf40..fe3e9dd 100644
--- a/fs/Makefile
+++ b/fs/Makefile
@@ -48,6 +48,8 @@ obj-$(CONFIG_COREDUMP) += coredump.o
obj-$(CONFIG_SYSCTL) += drop_caches.o

obj-$(CONFIG_FHANDLE) += fhandle.o
+obj-$(CONFIG_FS_RICHACL) += richacl.o
+richacl-y := richacl_base.o

obj-y += quota/

diff --git a/fs/richacl_base.c b/fs/richacl_base.c
new file mode 100644
index 0000000..6d9a073
--- /dev/null
+++ b/fs/richacl_base.c
@@ -0,0 +1,67 @@
+/*
+ * Copyright (C) 2006, 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#include <linux/sched.h>
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/richacl.h>
+
+MODULE_LICENSE("GPL");
+
+/**
+ * richacl_alloc - allocate a richacl
+ * @count: number of entries
+ */
+struct richacl *
+richacl_alloc(int count, gfp_t gfp)
+{
+ size_t size = sizeof(struct richacl) + count * sizeof(struct richace);
+ struct richacl *acl = kzalloc(size, gfp);
+
+ if (acl) {
+ atomic_set(&acl->a_refcount, 1);
+ acl->a_count = count;
+ }
+ return acl;
+}
+EXPORT_SYMBOL_GPL(richacl_alloc);
+
+/**
+ * richacl_clone - create a copy of a richacl
+ */
+struct richacl *
+richacl_clone(const struct richacl *acl, gfp_t gfp)
+{
+ int count = acl->a_count;
+ size_t size = sizeof(struct richacl) + count * sizeof(struct richace);
+ struct richacl *dup = kmalloc(size, gfp);
+
+ if (dup) {
+ memcpy(dup, acl, size);
+ atomic_set(&dup->a_refcount, 1);
+ }
+ return dup;
+}
+
+/**
+ * richace_copy - copy an acl entry
+ */
+void
+richace_copy(struct richace *to, const struct richace *from)
+{
+ memcpy(to, from, sizeof(struct richace));
+}
diff --git a/include/linux/richacl.h b/include/linux/richacl.h
new file mode 100644
index 0000000..bfa94bb
--- /dev/null
+++ b/include/linux/richacl.h
@@ -0,0 +1,256 @@
+/*
+ * Copyright (C) 2006, 2010 Novell, Inc.
+ * Copyright (C) 2015 Red Hat, Inc.
+ * Written by Andreas Gruenbacher <[email protected]>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef __RICHACL_H
+#define __RICHACL_H
+
+#define RICHACE_OWNER_SPECIAL_ID 0
+#define RICHACE_GROUP_SPECIAL_ID 1
+#define RICHACE_EVERYONE_SPECIAL_ID 2
+
+struct richace {
+ unsigned short e_type;
+ unsigned short e_flags;
+ unsigned int e_mask;
+ union {
+ kuid_t uid;
+ kgid_t gid;
+ unsigned int special;
+ } e_id;
+};
+
+struct richacl {
+ atomic_t a_refcount;
+ unsigned int a_owner_mask;
+ unsigned int a_group_mask;
+ unsigned int a_other_mask;
+ unsigned short a_count;
+ unsigned short a_flags;
+ struct richace a_entries[0];
+};
+
+#define richacl_for_each_entry(_ace, _acl) \
+ for (_ace = (_acl)->a_entries; \
+ _ace != (_acl)->a_entries + (_acl)->a_count; \
+ _ace++)
+
+#define richacl_for_each_entry_reverse(_ace, _acl) \
+ for (_ace = (_acl)->a_entries + (_acl)->a_count - 1; \
+ _ace != (_acl)->a_entries - 1; \
+ _ace--)
+
+/* a_flags values */
+#define RICHACL_WRITE_THROUGH 0x40
+#define RICHACL_MASKED 0x80
+
+#define RICHACL_VALID_FLAGS ( \
+ RICHACL_WRITE_THROUGH | \
+ RICHACL_MASKED)
+
+/* e_type values */
+#define RICHACE_ACCESS_ALLOWED_ACE_TYPE 0x0000
+#define RICHACE_ACCESS_DENIED_ACE_TYPE 0x0001
+
+/* e_flags bitflags */
+#define RICHACE_FILE_INHERIT_ACE 0x0001
+#define RICHACE_DIRECTORY_INHERIT_ACE 0x0002
+#define RICHACE_NO_PROPAGATE_INHERIT_ACE 0x0004
+#define RICHACE_INHERIT_ONLY_ACE 0x0008
+#define RICHACE_IDENTIFIER_GROUP 0x0040
+#define RICHACE_SPECIAL_WHO 0x4000
+
+#define RICHACE_VALID_FLAGS ( \
+ RICHACE_FILE_INHERIT_ACE | \
+ RICHACE_DIRECTORY_INHERIT_ACE | \
+ RICHACE_NO_PROPAGATE_INHERIT_ACE | \
+ RICHACE_INHERIT_ONLY_ACE | \
+ RICHACE_IDENTIFIER_GROUP | \
+ RICHACE_SPECIAL_WHO)
+
+#define RICHACE_INHERITANCE_FLAGS ( \
+ RICHACE_FILE_INHERIT_ACE | \
+ RICHACE_DIRECTORY_INHERIT_ACE | \
+ RICHACE_NO_PROPAGATE_INHERIT_ACE | \
+ RICHACE_INHERIT_ONLY_ACE )
+
+/* e_mask bitflags */
+#define RICHACE_READ_DATA 0x00000001
+#define RICHACE_LIST_DIRECTORY 0x00000001
+#define RICHACE_WRITE_DATA 0x00000002
+#define RICHACE_ADD_FILE 0x00000002
+#define RICHACE_APPEND_DATA 0x00000004
+#define RICHACE_ADD_SUBDIRECTORY 0x00000004
+#define RICHACE_READ_NAMED_ATTRS 0x00000008
+#define RICHACE_WRITE_NAMED_ATTRS 0x00000010
+#define RICHACE_EXECUTE 0x00000020
+#define RICHACE_DELETE_CHILD 0x00000040
+#define RICHACE_READ_ATTRIBUTES 0x00000080
+#define RICHACE_WRITE_ATTRIBUTES 0x00000100
+#define RICHACE_WRITE_RETENTION 0x00000200
+#define RICHACE_WRITE_RETENTION_HOLD 0x00000400
+#define RICHACE_DELETE 0x00010000
+#define RICHACE_READ_ACL 0x00020000
+#define RICHACE_WRITE_ACL 0x00040000
+#define RICHACE_WRITE_OWNER 0x00080000
+#define RICHACE_SYNCHRONIZE 0x00100000
+
+/* Valid RICHACE_* flags for directories and non-directories */
+#define RICHACE_VALID_MASK ( \
+ RICHACE_READ_DATA | RICHACE_LIST_DIRECTORY | \
+ RICHACE_WRITE_DATA | RICHACE_ADD_FILE | \
+ RICHACE_APPEND_DATA | RICHACE_ADD_SUBDIRECTORY | \
+ RICHACE_READ_NAMED_ATTRS | \
+ RICHACE_WRITE_NAMED_ATTRS | \
+ RICHACE_EXECUTE | \
+ RICHACE_DELETE_CHILD | \
+ RICHACE_READ_ATTRIBUTES | \
+ RICHACE_WRITE_ATTRIBUTES | \
+ RICHACE_WRITE_RETENTION | \
+ RICHACE_WRITE_RETENTION_HOLD | \
+ RICHACE_DELETE | \
+ RICHACE_READ_ACL | \
+ RICHACE_WRITE_ACL | \
+ RICHACE_WRITE_OWNER | \
+ RICHACE_SYNCHRONIZE)
+
+/**
+ * richacl_get - grab another reference to a richacl handle
+ */
+static inline struct richacl *
+richacl_get(struct richacl *acl)
+{
+ if (acl)
+ atomic_inc(&acl->a_refcount);
+ return acl;
+}
+
+/**
+ * richacl_put - free a richacl handle
+ */
+static inline void
+richacl_put(struct richacl *acl)
+{
+ if (acl && atomic_dec_and_test(&acl->a_refcount))
+ kfree(acl);
+}
+
+/**
+ * richace_is_owner - check if @ace is an OWNER@ entry
+ */
+static inline bool
+richace_is_owner(const struct richace *ace)
+{
+ return (ace->e_flags & RICHACE_SPECIAL_WHO) &&
+ ace->e_id.special == RICHACE_OWNER_SPECIAL_ID;
+}
+
+/**
+ * richace_is_group - check if @ace is a GROUP@ entry
+ */
+static inline bool
+richace_is_group(const struct richace *ace)
+{
+ return (ace->e_flags & RICHACE_SPECIAL_WHO) &&
+ ace->e_id.special == RICHACE_GROUP_SPECIAL_ID;
+}
+
+/**
+ * richace_is_everyone - check if @ace is an EVERYONE@ entry
+ */
+static inline bool
+richace_is_everyone(const struct richace *ace)
+{
+ return (ace->e_flags & RICHACE_SPECIAL_WHO) &&
+ ace->e_id.special == RICHACE_EVERYONE_SPECIAL_ID;
+}
+
+/**
+ * richace_is_unix_user - check if @ace applies to a specific user
+ */
+static inline bool
+richace_is_unix_user(const struct richace *ace)
+{
+ return !(ace->e_flags & RICHACE_SPECIAL_WHO) &&
+ !(ace->e_flags & RICHACE_IDENTIFIER_GROUP);
+}
+
+/**
+ * richace_is_unix_group - check if @ace applies to a specific group
+ */
+static inline bool
+richace_is_unix_group(const struct richace *ace)
+{
+ return !(ace->e_flags & RICHACE_SPECIAL_WHO) &&
+ (ace->e_flags & RICHACE_IDENTIFIER_GROUP);
+}
+
+/**
+ * richace_is_inherit_only - check if @ace is for inheritance only
+ *
+ * ACEs with the %RICHACE_INHERIT_ONLY_ACE flag set have no effect during
+ * permission checking.
+ */
+static inline bool
+richace_is_inherit_only(const struct richace *ace)
+{
+ return ace->e_flags & RICHACE_INHERIT_ONLY_ACE;
+}
+
+/**
+ * richace_is_inheritable - check if @ace is inheritable
+ */
+static inline bool
+richace_is_inheritable(const struct richace *ace)
+{
+ return ace->e_flags & (RICHACE_FILE_INHERIT_ACE |
+ RICHACE_DIRECTORY_INHERIT_ACE);
+}
+
+/**
+ * richace_is_allow - check if @ace is an %ALLOW type entry
+ */
+static inline bool
+richace_is_allow(const struct richace *ace)
+{
+ return ace->e_type == RICHACE_ACCESS_ALLOWED_ACE_TYPE;
+}
+
+/**
+ * richace_is_deny - check if @ace is a %DENY type entry
+ */
+static inline bool
+richace_is_deny(const struct richace *ace)
+{
+ return ace->e_type == RICHACE_ACCESS_DENIED_ACE_TYPE;
+}
+
+/**
+ * richace_is_same_identifier - are both identifiers the same?
+ */
+static inline bool
+richace_is_same_identifier(const struct richace *a, const struct richace *b)
+{
+ return !((a->e_flags ^ b->e_flags) &
+ (RICHACE_SPECIAL_WHO | RICHACE_IDENTIFIER_GROUP)) &&
+ !memcmp(&a->e_id, &b->e_id, sizeof(a->e_id));
+}
+
+extern struct richacl *richacl_alloc(int, gfp_t);
+extern struct richacl *richacl_clone(const struct richacl *, gfp_t);
+extern void richace_copy(struct richace *, const struct richace *);
+
+
+#endif /* __RICHACL_H */
--
2.4.3


2015-09-27 22:10:14

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 05/41] vfs: Add permission flags for setting file attributes

Richacls support permissions that allow to take ownership of a file,
change the file permissions, and set the file timestamps. Support that
by introducing new permission mask flags and by checking for those mask
flags in inode_change_ok().

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/attr.c | 79 +++++++++++++++++++++++++++++++++++++++++++++---------
include/linux/fs.h | 3 +++
2 files changed, 70 insertions(+), 12 deletions(-)

diff --git a/fs/attr.c b/fs/attr.c
index 328be71..85483e0 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -17,6 +17,65 @@
#include <linux/ima.h>

/**
+ * inode_extended_permission - permissions beyond read/write/execute
+ *
+ * Check for permissions that only richacls can currently grant.
+ */
+static int inode_extended_permission(struct inode *inode, int mask)
+{
+ if (!IS_RICHACL(inode))
+ return -EPERM;
+ return inode_permission(inode, mask);
+}
+
+static bool inode_uid_change_ok(struct inode *inode, kuid_t ia_uid)
+{
+ if (uid_eq(current_fsuid(), inode->i_uid) &&
+ uid_eq(ia_uid, inode->i_uid))
+ return true;
+ if (uid_eq(current_fsuid(), ia_uid) &&
+ inode_extended_permission(inode, MAY_TAKE_OWNERSHIP) == 0)
+ return true;
+ if (capable_wrt_inode_uidgid(inode, CAP_CHOWN))
+ return true;
+ return false;
+}
+
+static bool inode_gid_change_ok(struct inode *inode, kgid_t ia_gid)
+{
+ int in_group = in_group_p(ia_gid);
+ if (uid_eq(current_fsuid(), inode->i_uid) &&
+ (in_group || gid_eq(ia_gid, inode->i_gid)))
+ return true;
+ if (in_group && inode_extended_permission(inode, MAY_TAKE_OWNERSHIP) == 0)
+ return true;
+ if (capable_wrt_inode_uidgid(inode, CAP_CHOWN))
+ return true;
+ return false;
+}
+
+/**
+ * inode_owner_permitted_or_capable
+ *
+ * Check for permissions implicitly granted to the owner, like MAY_CHMOD or
+ * MAY_SET_TIMES. Equivalent to inode_owner_or_capable for file systems
+ * without support for those permissions.
+ */
+static bool inode_owner_permitted_or_capable(struct inode *inode, int mask)
+{
+ struct user_namespace *ns;
+
+ if (uid_eq(current_fsuid(), inode->i_uid))
+ return true;
+ if (inode_extended_permission(inode, mask) == 0)
+ return true;
+ ns = current_user_ns();
+ if (ns_capable(ns, CAP_FOWNER) && kuid_has_mapping(ns, inode->i_uid))
+ return true;
+ return false;
+}
+
+/**
* inode_change_ok - check if attribute changes to an inode are allowed
* @inode: inode to check
* @attr: attributes to change
@@ -47,22 +106,18 @@ int inode_change_ok(struct inode *inode, struct iattr *attr)
return 0;

/* Make sure a caller can chown. */
- if ((ia_valid & ATTR_UID) &&
- (!uid_eq(current_fsuid(), inode->i_uid) ||
- !uid_eq(attr->ia_uid, inode->i_uid)) &&
- !capable_wrt_inode_uidgid(inode, CAP_CHOWN))
- return -EPERM;
+ if (ia_valid & ATTR_UID)
+ if (!inode_uid_change_ok(inode, attr->ia_uid))
+ return -EPERM;

/* Make sure caller can chgrp. */
- if ((ia_valid & ATTR_GID) &&
- (!uid_eq(current_fsuid(), inode->i_uid) ||
- (!in_group_p(attr->ia_gid) && !gid_eq(attr->ia_gid, inode->i_gid))) &&
- !capable_wrt_inode_uidgid(inode, CAP_CHOWN))
- return -EPERM;
+ if (ia_valid & ATTR_GID)
+ if (!inode_gid_change_ok(inode, attr->ia_gid))
+ return -EPERM;

/* Make sure a caller can chmod. */
if (ia_valid & ATTR_MODE) {
- if (!inode_owner_or_capable(inode))
+ if (!inode_owner_permitted_or_capable(inode, MAY_CHMOD))
return -EPERM;
/* Also check the setgid bit! */
if (!in_group_p((ia_valid & ATTR_GID) ? attr->ia_gid :
@@ -73,7 +128,7 @@ int inode_change_ok(struct inode *inode, struct iattr *attr)

/* Check for setting the inode time. */
if (ia_valid & (ATTR_MTIME_SET | ATTR_ATIME_SET | ATTR_TIMES_SET)) {
- if (!inode_owner_or_capable(inode))
+ if (!inode_owner_permitted_or_capable(inode, MAY_SET_TIMES))
return -EPERM;
}

diff --git a/include/linux/fs.h b/include/linux/fs.h
index aab32c8..ba91a89 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -86,6 +86,9 @@ typedef void (dax_iodone_t)(struct buffer_head *bh_map, int uptodate);
#define MAY_CREATE_DIR 0x00000200
#define MAY_DELETE_CHILD 0x00000400
#define MAY_DELETE_SELF 0x00000800
+#define MAY_TAKE_OWNERSHIP 0x00001000
+#define MAY_CHMOD 0x00002000
+#define MAY_SET_TIMES 0x00004000

/*
* flags in file.f_mode. Note that FMODE_READ and FMODE_WRITE must correspond
--
2.4.3


2015-09-27 22:10:08

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 04/41] vfs: Make the inode passed to inode_change_ok non-const

We will need to call iop->permission and iop->get_acl from
inode_change_ok() for additional permission checks, and both take a
non-const inode.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/attr.c | 2 +-
include/linux/fs.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/attr.c b/fs/attr.c
index 6530ced..328be71 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -28,7 +28,7 @@
* Should be called as the first thing in ->setattr implementations,
* possibly after taking additional locks.
*/
-int inode_change_ok(const struct inode *inode, struct iattr *attr)
+int inode_change_ok(struct inode *inode, struct iattr *attr)
{
unsigned int ia_valid = attr->ia_valid;

diff --git a/include/linux/fs.h b/include/linux/fs.h
index 402acd7..aab32c8 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2871,7 +2871,7 @@ extern int buffer_migrate_page(struct address_space *,
#define buffer_migrate_page NULL
#endif

-extern int inode_change_ok(const struct inode *, struct iattr *);
+extern int inode_change_ok(struct inode *, struct iattr *);
extern int inode_newsize_ok(const struct inode *, loff_t offset);
extern void setattr_copy(struct inode *inode, const struct iattr *attr);

--
2.4.3


2015-09-27 22:10:03

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 03/41] vfs: Add MAY_DELETE_SELF and MAY_DELETE_CHILD permission flags

Normally, deleting a file requires MAY_WRITE access to the parent
directory. With richacls, a file may be deleted with MAY_DELETE_CHILD access
to the parent directory or with MAY_DELETE_SELF access to the file.

To support that, pass the MAY_DELETE_CHILD mask flag to inode_permission()
when checking for delete access inside a directory, and MAY_DELETE_SELF
when checking for delete access to a file itelf.

The MAY_DELETE_SELF permission overrides the sticky directory check.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/namei.c | 21 ++++++++++++---------
include/linux/fs.h | 2 ++
2 files changed, 14 insertions(+), 9 deletions(-)

diff --git a/fs/namei.c b/fs/namei.c
index 7c0f310..6865cdf 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -453,9 +453,9 @@ static int sb_permission(struct super_block *sb, struct inode *inode, int mask)
* this, letting us set arbitrary permissions for filesystem access without
* changing the "normal" UIDs which are used for other things.
*
- * MAY_WRITE must be set in @mask whenever MAY_APPEND, MAY_CREATE_FILE, or
- * MAY_CREATE_DIR are set. That way, file systems that don't support these
- * permissions will check for MAY_WRITE instead.
+ * MAY_WRITE must be set in @mask whenever MAY_APPEND, MAY_CREATE_FILE,
+ * MAY_CREATE_DIR, or MAY_DELETE_CHILD are set. That way, file systems that
+ * don't support these permissions will check for MAY_WRITE instead.
*/
int inode_permission(struct inode *inode, int mask)
{
@@ -2551,7 +2551,7 @@ static int may_delete(struct inode *dir, struct dentry *victim,
bool isdir, bool replace)
{
struct inode *inode = d_backing_inode(victim);
- int error, mask = MAY_WRITE | MAY_EXEC;
+ int error, mask = MAY_EXEC;

if (d_is_negative(victim))
return -ENOENT;
@@ -2561,15 +2561,18 @@ static int may_delete(struct inode *dir, struct dentry *victim,
audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);

if (replace)
- mask |= isdir ? MAY_CREATE_DIR : MAY_CREATE_FILE;
- error = inode_permission(dir, mask);
+ mask |= MAY_WRITE | (isdir ? MAY_CREATE_DIR : MAY_CREATE_FILE);
+ error = inode_permission(dir, mask | MAY_WRITE | MAY_DELETE_CHILD);
+ if (!error && check_sticky(dir, inode))
+ error = -EPERM;
+ if (error && IS_RICHACL(inode) &&
+ inode_permission(inode, MAY_DELETE_SELF) == 0)
+ error = 0;
if (error)
return error;
if (IS_APPEND(dir))
return -EPERM;
-
- if (check_sticky(dir, inode) || IS_APPEND(inode) ||
- IS_IMMUTABLE(inode) || IS_SWAPFILE(inode))
+ if (IS_APPEND(inode) || IS_IMMUTABLE(inode) || IS_SWAPFILE(inode))
return -EPERM;
if (isdir) {
if (!d_is_dir(victim))
diff --git a/include/linux/fs.h b/include/linux/fs.h
index d6e2330..402acd7 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -84,6 +84,8 @@ typedef void (dax_iodone_t)(struct buffer_head *bh_map, int uptodate);
#define MAY_NOT_BLOCK 0x00000080
#define MAY_CREATE_FILE 0x00000100
#define MAY_CREATE_DIR 0x00000200
+#define MAY_DELETE_CHILD 0x00000400
+#define MAY_DELETE_SELF 0x00000800

/*
* flags in file.f_mode. Note that FMODE_READ and FMODE_WRITE must correspond
--
2.4.3


2015-09-27 22:09:57

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 02/41] vfs: Add MAY_CREATE_FILE and MAY_CREATE_DIR permission flags

Richacls distinguish between creating non-directories and directories. To
support that, add an isdir parameter to may_create(). When checking
inode_permission() for create permission, pass in an additional
MAY_CREATE_FILE or MAY_CREATE_DIR mask flag.

To allow checking for delete *and* create access when replacing an existing
file via vfs_rename(), add a replace parameter to may_delete().

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/namei.c | 43 +++++++++++++++++++++++++------------------
include/linux/fs.h | 2 ++
2 files changed, 27 insertions(+), 18 deletions(-)

diff --git a/fs/namei.c b/fs/namei.c
index 48c2752..7c0f310 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -453,7 +453,9 @@ static int sb_permission(struct super_block *sb, struct inode *inode, int mask)
* this, letting us set arbitrary permissions for filesystem access without
* changing the "normal" UIDs which are used for other things.
*
- * When checking for MAY_APPEND, MAY_WRITE must also be set in @mask.
+ * MAY_WRITE must be set in @mask whenever MAY_APPEND, MAY_CREATE_FILE, or
+ * MAY_CREATE_DIR are set. That way, file systems that don't support these
+ * permissions will check for MAY_WRITE instead.
*/
int inode_permission(struct inode *inode, int mask)
{
@@ -2545,10 +2547,11 @@ EXPORT_SYMBOL(__check_sticky);
* 10. We don't allow removal of NFS sillyrenamed files; it's handled by
* nfs_async_unlink().
*/
-static int may_delete(struct inode *dir, struct dentry *victim, bool isdir)
+static int may_delete(struct inode *dir, struct dentry *victim,
+ bool isdir, bool replace)
{
struct inode *inode = d_backing_inode(victim);
- int error;
+ int error, mask = MAY_WRITE | MAY_EXEC;

if (d_is_negative(victim))
return -ENOENT;
@@ -2557,7 +2560,9 @@ static int may_delete(struct inode *dir, struct dentry *victim, bool isdir)
BUG_ON(victim->d_parent->d_inode != dir);
audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);

- error = inode_permission(dir, MAY_WRITE | MAY_EXEC);
+ if (replace)
+ mask |= isdir ? MAY_CREATE_DIR : MAY_CREATE_FILE;
+ error = inode_permission(dir, mask);
if (error)
return error;
if (IS_APPEND(dir))
@@ -2588,14 +2593,16 @@ static int may_delete(struct inode *dir, struct dentry *victim, bool isdir)
* 3. We should have write and exec permissions on dir
* 4. We can't do it if dir is immutable (done in permission())
*/
-static inline int may_create(struct inode *dir, struct dentry *child)
+static inline int may_create(struct inode *dir, struct dentry *child, bool isdir)
{
+ int mask = isdir ? MAY_CREATE_DIR : MAY_CREATE_FILE;
+
audit_inode_child(dir, child, AUDIT_TYPE_CHILD_CREATE);
if (child->d_inode)
return -EEXIST;
if (IS_DEADDIR(dir))
return -ENOENT;
- return inode_permission(dir, MAY_WRITE | MAY_EXEC);
+ return inode_permission(dir, MAY_WRITE | MAY_EXEC | mask);
}

/*
@@ -2645,7 +2652,7 @@ EXPORT_SYMBOL(unlock_rename);
int vfs_create(struct inode *dir, struct dentry *dentry, umode_t mode,
bool want_excl)
{
- int error = may_create(dir, dentry);
+ int error = may_create(dir, dentry, false);
if (error)
return error;

@@ -3490,7 +3497,7 @@ EXPORT_SYMBOL(user_path_create);

int vfs_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, dev_t dev)
{
- int error = may_create(dir, dentry);
+ int error = may_create(dir, dentry, false);

if (error)
return error;
@@ -3582,7 +3589,7 @@ SYSCALL_DEFINE3(mknod, const char __user *, filename, umode_t, mode, unsigned, d

int vfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
{
- int error = may_create(dir, dentry);
+ int error = may_create(dir, dentry, true);
unsigned max_links = dir->i_sb->s_max_links;

if (error)
@@ -3663,7 +3670,7 @@ EXPORT_SYMBOL(dentry_unhash);

int vfs_rmdir(struct inode *dir, struct dentry *dentry)
{
- int error = may_delete(dir, dentry, 1);
+ int error = may_delete(dir, dentry, true, false);

if (error)
return error;
@@ -3785,7 +3792,7 @@ SYSCALL_DEFINE1(rmdir, const char __user *, pathname)
int vfs_unlink(struct inode *dir, struct dentry *dentry, struct inode **delegated_inode)
{
struct inode *target = dentry->d_inode;
- int error = may_delete(dir, dentry, 0);
+ int error = may_delete(dir, dentry, false, false);

if (error)
return error;
@@ -3919,7 +3926,7 @@ SYSCALL_DEFINE1(unlink, const char __user *, pathname)

int vfs_symlink(struct inode *dir, struct dentry *dentry, const char *oldname)
{
- int error = may_create(dir, dentry);
+ int error = may_create(dir, dentry, false);

if (error)
return error;
@@ -4002,7 +4009,7 @@ int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_de
if (!inode)
return -ENOENT;

- error = may_create(dir, new_dentry);
+ error = may_create(dir, new_dentry, false);
if (error)
return error;

@@ -4190,19 +4197,19 @@ int vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
if (source == target)
return 0;

- error = may_delete(old_dir, old_dentry, is_dir);
+ error = may_delete(old_dir, old_dentry, is_dir, false);
if (error)
return error;

if (!target) {
- error = may_create(new_dir, new_dentry);
+ error = may_create(new_dir, new_dentry, is_dir);
} else {
new_is_dir = d_is_dir(new_dentry);

if (!(flags & RENAME_EXCHANGE))
- error = may_delete(new_dir, new_dentry, is_dir);
+ error = may_delete(new_dir, new_dentry, is_dir, true);
else
- error = may_delete(new_dir, new_dentry, new_is_dir);
+ error = may_delete(new_dir, new_dentry, new_is_dir, true);
}
if (error)
return error;
@@ -4465,7 +4472,7 @@ SYSCALL_DEFINE2(rename, const char __user *, oldname, const char __user *, newna

int vfs_whiteout(struct inode *dir, struct dentry *dentry)
{
- int error = may_create(dir, dentry);
+ int error = may_create(dir, dentry, false);
if (error)
return error;

diff --git a/include/linux/fs.h b/include/linux/fs.h
index 4efa435..d6e2330 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -82,6 +82,8 @@ typedef void (dax_iodone_t)(struct buffer_head *bh_map, int uptodate);
#define MAY_CHDIR 0x00000040
/* called from RCU mode, don't block */
#define MAY_NOT_BLOCK 0x00000080
+#define MAY_CREATE_FILE 0x00000100
+#define MAY_CREATE_DIR 0x00000200

/*
* flags in file.f_mode. Note that FMODE_READ and FMODE_WRITE must correspond
--
2.4.3


2015-09-27 22:09:51

by Andreas Gruenbacher

[permalink] [raw]
Subject: [PATCH v8 01/41] vfs: Add IS_ACL() and IS_RICHACL() tests

The vfs does not apply the umask for file systems that support acls. The
test used for this used to be called IS_POSIXACL(). Switch to a new
IS_ACL() test to check for either posix acls or richacls instead. Add a new
MS_RICHACL flag and IS_RICHACL() test for richacls alone. The IS_POSIXACL()
test is still needed by nfsd.

Signed-off-by: Andreas Gruenbacher <[email protected]>
Reviewed-by: J. Bruce Fields <[email protected]>
---
fs/Kconfig | 3 +++
fs/namei.c | 8 ++++----
include/linux/fs.h | 12 ++++++++++++
include/uapi/linux/fs.h | 3 ++-
4 files changed, 21 insertions(+), 5 deletions(-)

diff --git a/fs/Kconfig b/fs/Kconfig
index da3f32f..bff2879 100644
--- a/fs/Kconfig
+++ b/fs/Kconfig
@@ -56,6 +56,9 @@ endif # BLOCK
config FS_POSIX_ACL
def_bool n

+config FS_RICHACL
+ def_bool n
+
config EXPORTFS
tristate

diff --git a/fs/namei.c b/fs/namei.c
index 726d211..48c2752 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -2794,7 +2794,7 @@ static int atomic_open(struct nameidata *nd, struct dentry *dentry,
}

mode = op->mode;
- if ((open_flag & O_CREAT) && !IS_POSIXACL(dir))
+ if ((open_flag & O_CREAT) && !IS_ACL(dir))
mode &= ~current_umask();

excl = (open_flag & (O_EXCL | O_CREAT)) == (O_EXCL | O_CREAT);
@@ -2978,7 +2978,7 @@ static int lookup_open(struct nameidata *nd, struct path *path,
/* Negative dentry, just create the file */
if (!dentry->d_inode && (op->open_flag & O_CREAT)) {
umode_t mode = op->mode;
- if (!IS_POSIXACL(dir->d_inode))
+ if (!IS_ACL(dir->d_inode))
mode &= ~current_umask();
/*
* This write is needed to ensure that a
@@ -3549,7 +3549,7 @@ retry:
if (IS_ERR(dentry))
return PTR_ERR(dentry);

- if (!IS_POSIXACL(path.dentry->d_inode))
+ if (!IS_ACL(path.dentry->d_inode))
mode &= ~current_umask();
error = security_path_mknod(&path, dentry, mode, dev);
if (error)
@@ -3618,7 +3618,7 @@ retry:
if (IS_ERR(dentry))
return PTR_ERR(dentry);

- if (!IS_POSIXACL(path.dentry->d_inode))
+ if (!IS_ACL(path.dentry->d_inode))
mode &= ~current_umask();
error = security_path_mkdir(&path, dentry, mode);
if (!error)
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 72d8a84..4efa435 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1781,6 +1781,12 @@ struct super_operations {
#define IS_IMMUTABLE(inode) ((inode)->i_flags & S_IMMUTABLE)
#define IS_POSIXACL(inode) __IS_FLG(inode, MS_POSIXACL)

+#ifdef CONFIG_FS_RICHACL
+#define IS_RICHACL(inode) __IS_FLG(inode, MS_RICHACL)
+#else
+#define IS_RICHACL(inode) 0
+#endif
+
#define IS_DEADDIR(inode) ((inode)->i_flags & S_DEAD)
#define IS_NOCMTIME(inode) ((inode)->i_flags & S_NOCMTIME)
#define IS_SWAPFILE(inode) ((inode)->i_flags & S_SWAPFILE)
@@ -1794,6 +1800,12 @@ struct super_operations {
(inode)->i_rdev == WHITEOUT_DEV)

/*
+ * IS_ACL() tells the VFS to not apply the umask
+ * and use check_acl for acl permission checks when defined.
+ */
+#define IS_ACL(inode) __IS_FLG(inode, MS_POSIXACL | MS_RICHACL)
+
+/*
* Inode state bits. Protected by inode->i_lock
*
* Three bits determine the dirty state of the inode, I_DIRTY_SYNC,
diff --git a/include/uapi/linux/fs.h b/include/uapi/linux/fs.h
index 9b964a5..6ac6bc9 100644
--- a/include/uapi/linux/fs.h
+++ b/include/uapi/linux/fs.h
@@ -81,7 +81,7 @@ struct inodes_stat_t {
#define MS_VERBOSE 32768 /* War is peace. Verbosity is silence.
MS_VERBOSE is deprecated. */
#define MS_SILENT 32768
-#define MS_POSIXACL (1<<16) /* VFS does not apply the umask */
+#define MS_POSIXACL (1<<16) /* Supports POSIX ACLs */
#define MS_UNBINDABLE (1<<17) /* change to unbindable */
#define MS_PRIVATE (1<<18) /* change to private */
#define MS_SLAVE (1<<19) /* change to slave */
@@ -91,6 +91,7 @@ struct inodes_stat_t {
#define MS_I_VERSION (1<<23) /* Update inode I_version field */
#define MS_STRICTATIME (1<<24) /* Always perform atime updates */
#define MS_LAZYTIME (1<<25) /* Update the on-disk [acm]times lazily */
+#define MS_RICHACL (1<<26) /* Supports richacls */

/* These sb flags are internal to the kernel */
#define MS_NOSEC (1<<28)
--
2.4.3


2015-09-28 15:28:08

by J. Bruce Fields

[permalink] [raw]
Subject: Re: [PATCH v8 09/41] richacl: Update the file masks in chmod()

On Mon, Sep 28, 2015 at 12:09:00AM +0200, Andreas Gruenbacher wrote:
> Doing a chmod() sets the file mode, which includes the file permission
> bits. When a file has a richacl, the permissions that the richacl
> grants need to be limited to what the new file permission bits allow.
>
> This is done by setting the file masks in the richacl to what the file
> permission bits map to. The richacl access check algorithm takes the
> file masks into account, which ensures that the richacl cannot grant too
> many permissions.
>
> It is possible to explicitly add permissions to the file masks which go
> beyond what the file permission bits can grant (like the
> RICHACE_WRITE_ACL permission). The POSIX.1 standard calls this an
> alternate file access control mechanism. A subsequent chmod() would
> ensure that those permissions are disabled again.
>
> Signed-off-by: Andreas Gruenbacher <[email protected]>
> Reviewed-by: J. Bruce Fields <[email protected]>
> ---
> fs/richacl_base.c | 40 ++++++++++++++++++++++++++++++++++++++++
> include/linux/richacl.h | 1 +
> 2 files changed, 41 insertions(+)
>
> diff --git a/fs/richacl_base.c b/fs/richacl_base.c
> index fc544f7..6c69234 100644
> --- a/fs/richacl_base.c
> +++ b/fs/richacl_base.c
> @@ -339,3 +339,43 @@ restart:
> acl->a_flags &= ~(RICHACL_WRITE_THROUGH | RICHACL_MASKED);
> }
> EXPORT_SYMBOL_GPL(richacl_compute_max_masks);
> +
> +/**
> + * richacl_chmod - update the file masks to reflect the new mode
> + * @mode: new file permission bits including the file type
> + *
> + * Return a copy of @acl where the file masks have been replaced by the file
> + * masks corresponding to the file permission bits in @mode, or returns @acl
> + * itself if the file masks are already up to date. Takes over a reference
> + * to @acl.
> + */
> +struct richacl *
> +richacl_chmod(struct richacl *acl, mode_t mode)
> +{
> + unsigned int x = S_ISDIR(mode) ? 0 : RICHACE_DELETE_CHILD;
> + unsigned int owner_mask, group_mask, other_mask;
> + struct richacl *clone;
> +
> + owner_mask = richacl_mode_to_mask(mode >> 6) & ~x;
> + group_mask = richacl_mode_to_mask(mode >> 3) & ~x;
> + other_mask = richacl_mode_to_mask(mode) & ~x;
> +
> + if (acl->a_owner_mask == owner_mask &&
> + acl->a_group_mask == group_mask &&
> + acl->a_other_mask == other_mask &&
> + (acl->a_flags & (RICHACL_WRITE_THROUGH | RICHACL_MASKED)))

I think you meant to require both of those bits to be set.

--b.

> + return acl;
> +
> + clone = richacl_clone(acl, GFP_KERNEL);
> + richacl_put(acl);
> + if (!clone)
> + return ERR_PTR(-ENOMEM);
> +
> + clone->a_flags |= (RICHACL_WRITE_THROUGH | RICHACL_MASKED);
> + clone->a_owner_mask = owner_mask;
> + clone->a_group_mask = group_mask;
> + clone->a_other_mask = other_mask;
> +
> + return clone;
> +}
> +EXPORT_SYMBOL_GPL(richacl_chmod);
> diff --git a/include/linux/richacl.h b/include/linux/richacl.h
> index e81144a..e00f313 100644
> --- a/include/linux/richacl.h
> +++ b/include/linux/richacl.h
> @@ -298,5 +298,6 @@ extern int richacl_masks_to_mode(const struct richacl *);
> extern unsigned int richacl_mode_to_mask(mode_t);
> extern unsigned int richacl_want_to_mask(unsigned int);
> extern void richacl_compute_max_masks(struct richacl *);
> +extern struct richacl *richacl_chmod(struct richacl *, mode_t);
>
> #endif /* __RICHACL_H */
> --
> 2.4.3

2015-09-28 16:08:56

by J. Bruce Fields

[permalink] [raw]
Subject: Re: [PATCH v8 10/41] richacl: Permission check algorithm

On Mon, Sep 28, 2015 at 12:09:01AM +0200, Andreas Gruenbacher wrote:
> A richacl roughly grants a requested access if the NFSv4 acl in the
> richacl grants the requested permissions according to the NFSv4
> permission check algorithm and the file mask that applies to the process
> includes the requested permissions.
>
> Signed-off-by: Andreas Gruenbacher <[email protected]>
> Reviewed-by: "J. Bruce Fields" <[email protected]>
> ---
> fs/Makefile | 2 +-
> fs/richacl_inode.c | 148 ++++++++++++++++++++++++++++++++++++++++++++++++
> include/linux/richacl.h | 3 +
> 3 files changed, 152 insertions(+), 1 deletion(-)
> create mode 100644 fs/richacl_inode.c
>
> diff --git a/fs/Makefile b/fs/Makefile
> index fe3e9dd..ec665fd 100644
> --- a/fs/Makefile
> +++ b/fs/Makefile
> @@ -49,7 +49,7 @@ obj-$(CONFIG_SYSCTL) += drop_caches.o
>
> obj-$(CONFIG_FHANDLE) += fhandle.o
> obj-$(CONFIG_FS_RICHACL) += richacl.o
> -richacl-y := richacl_base.o
> +richacl-y := richacl_base.o richacl_inode.o
>
> obj-y += quota/
>
> diff --git a/fs/richacl_inode.c b/fs/richacl_inode.c
> new file mode 100644
> index 0000000..54e899d
> --- /dev/null
> +++ b/fs/richacl_inode.c
> @@ -0,0 +1,148 @@
> +/*
> + * Copyright (C) 2010 Novell, Inc.
> + * Copyright (C) 2015 Red Hat, Inc.
> + * Written by Andreas Gruenbacher <[email protected]>
> + *
> + * This program is free software; you can redistribute it and/or modify it
> + * under the terms of the GNU General Public License as published by the
> + * Free Software Foundation; either version 2, or (at your option) any
> + * later version.
> + *
> + * This program is distributed in the hope that it will be useful, but
> + * WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * General Public License for more details.
> + */
> +
> +#include <linux/sched.h>
> +#include <linux/module.h>
> +#include <linux/fs.h>
> +#include <linux/slab.h>
> +#include <linux/richacl.h>
> +
> +/**
> + * richacl_permission - richacl permission check algorithm
> + * @inode: inode to check
> + * @acl: rich acl of the inode
> + * @want: requested access (MAY_* flags)
> + *
> + * Checks if the current process is granted @mask flags in @acl.
> + */
> +int
> +richacl_permission(struct inode *inode, const struct richacl *acl,
> + int want)
> +{
> + const struct richace *ace;
> + unsigned int mask = richacl_want_to_mask(want);
> + unsigned int requested = mask, denied = 0;
> + int in_owning_group = in_group_p(inode->i_gid);
> + int in_owner_or_group_class = in_owning_group;
> +
> + /*
> + * A process is
> + * - in the owner file class if it owns the file,
> + * - in the group file class if it is in the file's owning group or
> + * it matches any of the user or group entries, and
> + * - in the other file class otherwise.
> + * The file class is only relevant for determining which file mask to
> + * apply, which only happens for masked acls.
> + */
> + if (acl->a_flags & RICHACL_MASKED) {
> + if ((acl->a_flags & RICHACL_WRITE_THROUGH) &&
> + uid_eq(current_fsuid(), inode->i_uid)) {
> + denied = requested & ~acl->a_owner_mask;
> + goto out;
> + }
> + } else {
> + /*
> + * When the acl is not masked, there is no need to determine if
> + * the process is in the group class and we can break out
> + * earlier of the loop below.
> + */
> + in_owner_or_group_class = 1;
> + }
> +
> + /*
> + * Check if the acl grants the requested access and determine which
> + * file class the process is in.
> + */
> + richacl_for_each_entry(ace, acl) {
> + unsigned int ace_mask = ace->e_mask;
> +
> + if (richace_is_inherit_only(ace))
> + continue;
> + if (richace_is_owner(ace)) {
> + if (!uid_eq(current_fsuid(), inode->i_uid))
> + continue;
> + goto entry_matches_owner;
> + } else if (richace_is_group(ace)) {
> + if (!in_owning_group)
> + continue;
> + } else if (richace_is_unix_user(ace)) {
> + if (!uid_eq(current_fsuid(), ace->e_id.uid))
> + continue;
> + goto entry_matches_owner;
> + } else if (richace_is_unix_group(ace)) {
> + if (!in_group_p(ace->e_id.gid))
> + continue;
> + } else
> + goto entry_matches_everyone;
> +
> + /*
> + * Apply the group file mask to entries other than owner@ and
> + * everyone@ or user entries matching the owner.

The above also skips the following group_mask application on any unix
group. But that's OK, since a non-owner matching a unix group will be
placed into the group class, so the RICHACL_MASKED clause after the loop
below will apply the group mask. That logic is a little subtle, but OK.

--b.

> This ensures
> + * that we grant the same permissions as the acl computed by
> + * richacl_apply_masks().
> + *
> + * Without this restriction, the following richacl would grant
> + * rw access to processes which are both the owner and in the
> + * owning group, but not to other users in the owning group,
> + * which could not be represented without masks:
> + *
> + * owner:rw::mask
> + * group@:rw::allow
> + */
> + if ((acl->a_flags & RICHACL_MASKED) && richace_is_allow(ace))
> + ace_mask &= acl->a_group_mask;
> +
> +entry_matches_owner:
> + /* The process is in the owner or group file class. */
> + in_owner_or_group_class = 1;
> +
> +entry_matches_everyone:
> + /* Check which mask flags the ACE allows or denies. */
> + if (richace_is_deny(ace))
> + denied |= ace_mask & mask;
> + mask &= ~ace_mask;
> +
> + /*
> + * Keep going until we know which file class
> + * the process is in.
> + */
> + if (!mask && in_owner_or_group_class)
> + break;
> + }
> + denied |= mask;
> +
> + if (acl->a_flags & RICHACL_MASKED) {
> + /*
> + * The file class a process is in determines which file mask
> + * applies. Check if that file mask also grants the requested
> + * access.
> + */
> + if (uid_eq(current_fsuid(), inode->i_uid))
> + denied |= requested & ~acl->a_owner_mask;
> + else if (in_owner_or_group_class)
> + denied |= requested & ~acl->a_group_mask;
> + else {
> + if (acl->a_flags & RICHACL_WRITE_THROUGH)
> + denied = requested & ~acl->a_other_mask;
> + else
> + denied |= requested & ~acl->a_other_mask;
> + }
> + }
> +
> +out:
> + return denied ? -EACCES : 0;
> +}
> +EXPORT_SYMBOL_GPL(richacl_permission);
> diff --git a/include/linux/richacl.h b/include/linux/richacl.h
> index e00f313..9768eeb 100644
> --- a/include/linux/richacl.h
> +++ b/include/linux/richacl.h
> @@ -300,4 +300,7 @@ extern unsigned int richacl_want_to_mask(unsigned int);
> extern void richacl_compute_max_masks(struct richacl *);
> extern struct richacl *richacl_chmod(struct richacl *, mode_t);
>
> +/* richacl_inode.c */
> +extern int richacl_permission(struct inode *, const struct richacl *, int);
> +
> #endif /* __RICHACL_H */
> --
> 2.4.3

2015-09-28 16:25:25

by Andreas Grünbacher

[permalink] [raw]
Subject: Re: [PATCH v8 10/41] richacl: Permission check algorithm

2015-09-28 18:08 GMT+02:00 J. Bruce Fields <[email protected]>:
> On Mon, Sep 28, 2015 at 12:09:01AM +0200, Andreas Gruenbacher wrote:
>> + /*
>> + * Check if the acl grants the requested access and determine which
>> + * file class the process is in.
>> + */
>> + richacl_for_each_entry(ace, acl) {
>> + unsigned int ace_mask = ace->e_mask;
>> +
>> + if (richace_is_inherit_only(ace))
>> + continue;
>> + if (richace_is_owner(ace)) {
>> + if (!uid_eq(current_fsuid(), inode->i_uid))
>> + continue;
>> + goto entry_matches_owner;
>> + } else if (richace_is_group(ace)) {
>> + if (!in_owning_group)
>> + continue;
>> + } else if (richace_is_unix_user(ace)) {
>> + if (!uid_eq(current_fsuid(), ace->e_id.uid))
>> + continue;
>> + goto entry_matches_owner;
>> + } else if (richace_is_unix_group(ace)) {
>> + if (!in_group_p(ace->e_id.gid))
>> + continue;
>> + } else
>> + goto entry_matches_everyone;
>> +
>> + /*
>> + * Apply the group file mask to entries other than owner@ and
>> + * everyone@ or user entries matching the owner.
>
> The above also skips the following group_mask application on any unix
> group.

Really? How does it do that?

Thanks,
Andreas

2015-09-28 16:29:07

by J. Bruce Fields

[permalink] [raw]
Subject: Re: [PATCH v8 10/41] richacl: Permission check algorithm

On Mon, Sep 28, 2015 at 06:25:23PM +0200, Andreas Grünbacher wrote:
> 2015-09-28 18:08 GMT+02:00 J. Bruce Fields <[email protected]>:
> > On Mon, Sep 28, 2015 at 12:09:01AM +0200, Andreas Gruenbacher wrote:
> >> + /*
> >> + * Check if the acl grants the requested access and determine which
> >> + * file class the process is in.
> >> + */
> >> + richacl_for_each_entry(ace, acl) {
> >> + unsigned int ace_mask = ace->e_mask;
> >> +
> >> + if (richace_is_inherit_only(ace))
> >> + continue;
> >> + if (richace_is_owner(ace)) {
> >> + if (!uid_eq(current_fsuid(), inode->i_uid))
> >> + continue;
> >> + goto entry_matches_owner;
> >> + } else if (richace_is_group(ace)) {
> >> + if (!in_owning_group)
> >> + continue;
> >> + } else if (richace_is_unix_user(ace)) {
> >> + if (!uid_eq(current_fsuid(), ace->e_id.uid))
> >> + continue;
> >> + goto entry_matches_owner;
> >> + } else if (richace_is_unix_group(ace)) {
> >> + if (!in_group_p(ace->e_id.gid))
> >> + continue;
> >> + } else
> >> + goto entry_matches_everyone;
> >> +
> >> + /*
> >> + * Apply the group file mask to entries other than owner@ and
> >> + * everyone@ or user entries matching the owner.
> >
> > The above also skips the following group_mask application on any unix
> > group.
>
> Really? How does it do that?

Sorry, I meant "unix user", not "unix group"!

--b.

2015-09-28 16:35:44

by J. Bruce Fields

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
> here's another update of the richacl patch queue. At this stage, I would
> like to ask for final feedback so that the core and ext4 code (patches
> 1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
> then go through the respective maintainer trees.

I've been over the core richacl and nfsd parts very carefully, and they
definitely look ready to me.

> Changes since the last posting (https://lwn.net/Articles/656704/):
>
> * The MAY_DELETE_SELF permission now also overrides the sticky
> directory checks.
>
> * Fix the permission check algorithm to apply the owner mask instead
> of the group mask to user entries matching the current owner. That way,
> the owner will retain the permissions in those entries when creating
> objects with create mode 0700 and similar. (A chmod to mode 0700 already
> creates an owner@:rwpx::allow ace, which was hiding this bug.)
>
> * Fix richacl_apply_masks to properly insert deny aces when raising the
> permissions of the other class. The bug could be triggered by
> chmod'ing a group@:r::allow acl to mode 0077, for example.
>
> * Various cleanups and improvements to comments.
>
>
> The complete patch queue is available here:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/agruen/linux-richacl.git \
> richacl-2015-09-28
>
>
> The richacl user-space utilitites and test suite are available here:
>
> https://github.com/andreas-gruenbacher/richacl/
>
>
> Open issues in nfs:
>
> * When a user or group name cannot be mapped, nfs's idmapper always maps it
> to nobody. That's good enough for mapping the file owner and owning
> group, but not for identifiers in acls. For now, to get the nfs richacl
> support somewhat working, I'm explicitly checking if mapping has resulted
> in uid/gid 99 in the kernel.
>
> * When the nfs server replies with NFS4ERR_BADNAME for any user or group
> name lookup, the client will stop sending numeric uids and gids to the
> server even when the lookup wasn't numeric. From then on, the client
> will translate uids and gids that have no mapping to the string "nobody",
> and the server will reject them. This problem is not specific to acls.

Do you have fixes in mind for these two issues?

--b.


>
> Thanks,
> Andreas
>
> Andreas Gruenbacher (39):
> vfs: Add IS_ACL() and IS_RICHACL() tests
> vfs: Add MAY_CREATE_FILE and MAY_CREATE_DIR permission flags
> vfs: Add MAY_DELETE_SELF and MAY_DELETE_CHILD permission flags
> vfs: Make the inode passed to inode_change_ok non-const
> vfs: Add permission flags for setting file attributes
> richacl: In-memory representation and helper functions
> richacl: Permission mapping functions
> richacl: Compute maximum file masks from an acl
> richacl: Update the file masks in chmod()
> richacl: Permission check algorithm
> vfs: Cache base_acl objects in inodes
> vfs: Cache richacl in struct inode
> richacl: Check if an acl is equivalent to a file mode
> richacl: Create-time inheritance
> richacl: Automatic Inheritance
> richacl: xattr mapping functions
> vfs: Add richacl permission checking
> richacl: acl editing helper functions
> richacl: Move everyone@ aces down the acl
> richacl: Propagate everyone@ permissions to other aces
> richacl: Set the owner permissions to the owner mask
> richacl: Set the other permissions to the other mask
> richacl: Isolate the owner and group classes
> richacl: Apply the file masks to a richacl
> richacl: Create richacl from mode values
> nfsd: Keep list of acls to dispose of in compoundargs
> nfsd: Use richacls as internal acl representation
> nfsd: Add richacl support
> nfsd: Add support for the v4.1 dacl attribute
> nfsd: Add support for the MAY_CREATE_{FILE,DIR} permissions
> richacl: Add support for unmapped identifiers
> ext4: Don't allow unmapped identifiers in richacls
> sunrpc: Allow to demand-allocate pages to encode into
> sunrpc: Add xdr_init_encode_pages
> nfs: Fix GETATTR bitmap verification
> nfs: Remove unused xdr page offsets in getacl/setacl arguments
> nfs: Add richacl support
> nfs: Add support for the v4.1 dacl attribute
> richacl: uapi header split
>
> Aneesh Kumar K.V (2):
> ext4: Add richacl support
> ext4: Add richacl feature flag
>
> drivers/staging/lustre/lustre/llite/llite_lib.c | 2 +-
> fs/Kconfig | 9 +
> fs/Makefile | 3 +
> fs/attr.c | 81 ++-
> fs/ext4/Kconfig | 15 +
> fs/ext4/Makefile | 1 +
> fs/ext4/acl.c | 6 +-
> fs/ext4/acl.h | 12 +-
> fs/ext4/ext4.h | 6 +-
> fs/ext4/file.c | 6 +-
> fs/ext4/ialloc.c | 7 +-
> fs/ext4/inode.c | 10 +-
> fs/ext4/namei.c | 11 +-
> fs/ext4/richacl.c | 218 ++++++
> fs/ext4/richacl.h | 47 ++
> fs/ext4/super.c | 42 +-
> fs/ext4/xattr.c | 6 +
> fs/ext4/xattr.h | 1 +
> fs/f2fs/acl.c | 4 +-
> fs/inode.c | 15 +-
> fs/jffs2/acl.c | 6 +-
> fs/namei.c | 111 ++-
> fs/nfs/inode.c | 3 -
> fs/nfs/nfs4proc.c | 701 +++++++++++++-----
> fs/nfs/nfs4xdr.c | 257 ++++++-
> fs/nfs/super.c | 4 +-
> fs/nfs_common/Makefile | 1 +
> fs/nfs_common/nfs4acl.c | 44 ++
> fs/nfsd/Kconfig | 1 +
> fs/nfsd/acl.h | 23 +-
> fs/nfsd/nfs4acl.c | 482 +++++++------
> fs/nfsd/nfs4proc.c | 25 +-
> fs/nfsd/nfs4xdr.c | 268 ++++---
> fs/nfsd/nfsd.h | 6 +-
> fs/nfsd/nfsfh.c | 8 +-
> fs/nfsd/vfs.c | 28 +-
> fs/nfsd/vfs.h | 17 +-
> fs/nfsd/xdr4.h | 12 +-
> fs/posix_acl.c | 26 +-
> fs/richacl_base.c | 682 ++++++++++++++++++
> fs/richacl_compat.c | 915 ++++++++++++++++++++++++
> fs/richacl_inode.c | 297 ++++++++
> fs/richacl_xattr.c | 267 +++++++
> fs/xattr.c | 34 +-
> include/linux/fs.h | 50 +-
> include/linux/nfs4.h | 24 +-
> include/linux/nfs4acl.h | 7 +
> include/linux/nfs_fs.h | 1 -
> include/linux/nfs_fs_sb.h | 2 +
> include/linux/nfs_xdr.h | 13 +-
> include/linux/posix_acl.h | 12 +-
> include/linux/richacl.h | 275 +++++++
> include/linux/richacl_compat.h | 40 ++
> include/linux/richacl_xattr.h | 47 ++
> include/linux/sunrpc/xdr.h | 2 +
> include/uapi/linux/Kbuild | 2 +
> include/uapi/linux/fs.h | 3 +-
> include/uapi/linux/nfs4.h | 3 +-
> include/uapi/linux/richacl.h | 111 +++
> include/uapi/linux/richacl_xattr.h | 43 ++
> include/uapi/linux/xattr.h | 2 +
> net/sunrpc/xdr.c | 34 +
> 62 files changed, 4659 insertions(+), 732 deletions(-)
> create mode 100644 fs/ext4/richacl.c
> create mode 100644 fs/ext4/richacl.h
> create mode 100644 fs/nfs_common/nfs4acl.c
> create mode 100644 fs/richacl_base.c
> create mode 100644 fs/richacl_compat.c
> create mode 100644 fs/richacl_inode.c
> create mode 100644 fs/richacl_xattr.c
> create mode 100644 include/linux/nfs4acl.h
> create mode 100644 include/linux/richacl.h
> create mode 100644 include/linux/richacl_compat.h
> create mode 100644 include/linux/richacl_xattr.h
> create mode 100644 include/uapi/linux/richacl.h
> create mode 100644 include/uapi/linux/richacl_xattr.h
>
> --
> 2.4.3
>
>
> Andreas Gruenbacher (39):
> vfs: Add IS_ACL() and IS_RICHACL() tests
> vfs: Add MAY_CREATE_FILE and MAY_CREATE_DIR permission flags
> vfs: Add MAY_DELETE_SELF and MAY_DELETE_CHILD permission flags
> vfs: Make the inode passed to inode_change_ok non-const
> vfs: Add permission flags for setting file attributes
> richacl: In-memory representation and helper functions
> richacl: Permission mapping functions
> richacl: Compute maximum file masks from an acl
> richacl: Update the file masks in chmod()
> richacl: Permission check algorithm
> vfs: Cache base_acl objects in inodes
> vfs: Cache richacl in struct inode
> richacl: Check if an acl is equivalent to a file mode
> richacl: Create-time inheritance
> richacl: Automatic Inheritance
> richacl: xattr mapping functions
> vfs: Add richacl permission checking
> richacl: acl editing helper functions
> richacl: Move everyone@ aces down the acl
> richacl: Propagate everyone@ permissions to other aces
> richacl: Set the owner permissions to the owner mask
> richacl: Set the other permissions to the other mask
> richacl: Isolate the owner and group classes
> richacl: Apply the file masks to a richacl
> richacl: Create richacl from mode values
> nfsd: Keep list of acls to dispose of in compoundargs
> nfsd: Use richacls as internal acl representation
> nfsd: Add richacl support
> nfsd: Add support for the v4.1 dacl attribute
> nfsd: Add support for the MAY_CREATE_{FILE,DIR} permissions
> richacl: Add support for unmapped identifiers
> ext4: Don't allow unmapped identifiers in richacls
> sunrpc: Allow to demand-allocate pages to encode into
> sunrpc: Add xdr_init_encode_pages
> nfs: Fix GETATTR bitmap verification
> nfs: Remove unused xdr page offsets in getacl/setacl arguments
> nfs: Add richacl support
> nfs: Add support for the v4.1 dacl attribute
> richacl: uapi header split
>
> Aneesh Kumar K.V (2):
> ext4: Add richacl support
> ext4: Add richacl feature flag
>
> drivers/staging/lustre/lustre/llite/llite_lib.c | 2 +-
> fs/Kconfig | 9 +
> fs/Makefile | 3 +
> fs/attr.c | 81 ++-
> fs/ext4/Kconfig | 15 +
> fs/ext4/Makefile | 1 +
> fs/ext4/acl.c | 6 +-
> fs/ext4/acl.h | 12 +-
> fs/ext4/ext4.h | 6 +-
> fs/ext4/file.c | 6 +-
> fs/ext4/ialloc.c | 7 +-
> fs/ext4/inode.c | 10 +-
> fs/ext4/namei.c | 11 +-
> fs/ext4/richacl.c | 218 ++++++
> fs/ext4/richacl.h | 47 ++
> fs/ext4/super.c | 42 +-
> fs/ext4/xattr.c | 6 +
> fs/ext4/xattr.h | 1 +
> fs/f2fs/acl.c | 4 +-
> fs/inode.c | 15 +-
> fs/jffs2/acl.c | 6 +-
> fs/namei.c | 111 ++-
> fs/nfs/inode.c | 3 -
> fs/nfs/nfs4proc.c | 701 +++++++++++++-----
> fs/nfs/nfs4xdr.c | 257 ++++++-
> fs/nfs/super.c | 4 +-
> fs/nfs_common/Makefile | 1 +
> fs/nfs_common/nfs4acl.c | 44 ++
> fs/nfsd/Kconfig | 1 +
> fs/nfsd/acl.h | 23 +-
> fs/nfsd/nfs4acl.c | 482 +++++++------
> fs/nfsd/nfs4proc.c | 25 +-
> fs/nfsd/nfs4xdr.c | 268 ++++---
> fs/nfsd/nfsd.h | 6 +-
> fs/nfsd/nfsfh.c | 8 +-
> fs/nfsd/vfs.c | 28 +-
> fs/nfsd/vfs.h | 17 +-
> fs/nfsd/xdr4.h | 12 +-
> fs/posix_acl.c | 26 +-
> fs/richacl_base.c | 682 ++++++++++++++++++
> fs/richacl_compat.c | 915 ++++++++++++++++++++++++
> fs/richacl_inode.c | 297 ++++++++
> fs/richacl_xattr.c | 267 +++++++
> fs/xattr.c | 34 +-
> include/linux/fs.h | 50 +-
> include/linux/nfs4.h | 24 +-
> include/linux/nfs4acl.h | 7 +
> include/linux/nfs_fs.h | 1 -
> include/linux/nfs_fs_sb.h | 2 +
> include/linux/nfs_xdr.h | 13 +-
> include/linux/posix_acl.h | 12 +-
> include/linux/richacl.h | 275 +++++++
> include/linux/richacl_compat.h | 40 ++
> include/linux/richacl_xattr.h | 47 ++
> include/linux/sunrpc/xdr.h | 2 +
> include/uapi/linux/Kbuild | 2 +
> include/uapi/linux/fs.h | 3 +-
> include/uapi/linux/nfs4.h | 3 +-
> include/uapi/linux/richacl.h | 111 +++
> include/uapi/linux/richacl_xattr.h | 43 ++
> include/uapi/linux/xattr.h | 2 +
> net/sunrpc/xdr.c | 34 +
> 62 files changed, 4659 insertions(+), 732 deletions(-)
> create mode 100644 fs/ext4/richacl.c
> create mode 100644 fs/ext4/richacl.h
> create mode 100644 fs/nfs_common/nfs4acl.c
> create mode 100644 fs/richacl_base.c
> create mode 100644 fs/richacl_compat.c
> create mode 100644 fs/richacl_inode.c
> create mode 100644 fs/richacl_xattr.c
> create mode 100644 include/linux/nfs4acl.h
> create mode 100644 include/linux/richacl.h
> create mode 100644 include/linux/richacl_compat.h
> create mode 100644 include/linux/richacl_xattr.h
> create mode 100644 include/uapi/linux/richacl.h
> create mode 100644 include/uapi/linux/richacl_xattr.h
>
> --
> 2.4.3

2015-09-28 16:51:49

by Andreas Grünbacher

[permalink] [raw]
Subject: Re: [PATCH v8 10/41] richacl: Permission check algorithm

2015-09-28 18:29 GMT+02:00 J. Bruce Fields <[email protected]>:
> On Mon, Sep 28, 2015 at 06:25:23PM +0200, Andreas Grünbacher wrote:
>> 2015-09-28 18:08 GMT+02:00 J. Bruce Fields <[email protected]>:
>> > The above also skips the following group_mask application on any unix
>> > group.
>>
>> Really? How does it do that?
>
> Sorry, I meant "unix user", not "unix group"!

Indeed, that's a bit tricky. Probably worth changing if just for clarity:

- goto entry_matches_owner;
+ if (uid_eq(current_fsuid(), inode->i_uid))
+ goto entry_matches_owner;

Thanks,
Andreas

2015-09-28 17:10:08

by Andreas Grünbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

2015-09-28 18:35 GMT+02:00 J. Bruce Fields <[email protected]>:
> On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
>> here's another update of the richacl patch queue. At this stage, I would
>> like to ask for final feedback so that the core and ext4 code (patches
>> 1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
>> then go through the respective maintainer trees.
>
> I've been over the core richacl and nfsd parts very carefully, and they
> definitely look ready to me.

Thanks a lot for all that work, by the way.

>> Changes since the last posting (https://lwn.net/Articles/656704/):
>>
>> * The MAY_DELETE_SELF permission now also overrides the sticky
>> directory checks.
>>
>> * Fix the permission check algorithm to apply the owner mask instead
>> of the group mask to user entries matching the current owner. That way,
>> the owner will retain the permissions in those entries when creating
>> objects with create mode 0700 and similar. (A chmod to mode 0700 already
>> creates an owner@:rwpx::allow ace, which was hiding this bug.)
>>
>> * Fix richacl_apply_masks to properly insert deny aces when raising the
>> permissions of the other class. The bug could be triggered by
>> chmod'ing a group@:r::allow acl to mode 0077, for example.
>>
>> * Various cleanups and improvements to comments.
>>
>>
>> The complete patch queue is available here:
>>
>> git://git.kernel.org/pub/scm/linux/kernel/git/agruen/linux-richacl.git \
>> richacl-2015-09-28
>>
>>
>> The richacl user-space utilitites and test suite are available here:
>>
>> https://github.com/andreas-gruenbacher/richacl/
>>
>>
>> Open issues in nfs:
>>
>> * When a user or group name cannot be mapped, nfs's idmapper always maps it
>> to nobody. That's good enough for mapping the file owner and owning
>> group, but not for identifiers in acls. For now, to get the nfs richacl
>> support somewhat working, I'm explicitly checking if mapping has resulted
>> in uid/gid 99 in the kernel.
>>
>> * When the nfs server replies with NFS4ERR_BADNAME for any user or group
>> name lookup, the client will stop sending numeric uids and gids to the
>> server even when the lookup wasn't numeric. From then on, the client
>> will translate uids and gids that have no mapping to the string "nobody",
>> and the server will reject them. This problem is not specific to acls.
>
> Do you have fixes in mind for these two issues?

I'm not sure how to best fix the idmapper problem, with backwards
compatibility and all. The second problem shouldn't be too hard to
fix.

Thanks,
Andreas

2015-09-28 17:46:52

by J. Bruce Fields

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Mon, Sep 28, 2015 at 07:10:06PM +0200, Andreas Grünbacher wrote:
> 2015-09-28 18:35 GMT+02:00 J. Bruce Fields <[email protected]>:
> > On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
> >> Open issues in nfs:
> >>
> >> * When a user or group name cannot be mapped, nfs's idmapper always maps it
> >> to nobody. That's good enough for mapping the file owner and owning
> >> group, but not for identifiers in acls. For now, to get the nfs richacl
> >> support somewhat working, I'm explicitly checking if mapping has resulted
> >> in uid/gid 99 in the kernel.
> >>
> >> * When the nfs server replies with NFS4ERR_BADNAME for any user or group
> >> name lookup, the client will stop sending numeric uids and gids to the
> >> server even when the lookup wasn't numeric. From then on, the client
> >> will translate uids and gids that have no mapping to the string "nobody",
> >> and the server will reject them. This problem is not specific to acls.
> >
> > Do you have fixes in mind for these two issues?
>
> I'm not sure how to best fix the idmapper problem, with backwards
> compatibility and all.

I haven't looked at the current nfsidmap interface.... So it's
completely lacking any way to communicate failure?

> The second problem shouldn't be too hard to fix.

Is it enough to turn off the failover in the case there's no possibility
it could have been caused by a numeric id?

If any user can set ACLs with arbitrary strings as names, then we'd be
giving any user unprivileged user the ability to turn off numeric
idmapping, so I think we need to fix that.

--b.

2015-09-28 23:42:06

by Andreas Gruenbacher

[permalink] [raw]
Subject: Re: [PATCH v8 09/41] richacl: Update the file masks in chmod()

2015-09-28 17:28 GMT+02:00 J. Bruce Fields <[email protected]>:
> On Mon, Sep 28, 2015 at 12:09:00AM +0200, Andreas Gruenbacher wrote:
>> + (acl->a_flags & (RICHACL_WRITE_THROUGH | RICHACL_MASKED)))
>
> I think you meant to require both of those bits to be set.

Yes, indeed.

Thanks,
Andreas

2015-09-29 14:54:39

by Andreas Grünbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

2015-09-28 19:46 GMT+02:00 J. Bruce Fields <[email protected]>:
> On Mon, Sep 28, 2015 at 07:10:06PM +0200, Andreas Grünbacher wrote:
>> 2015-09-28 18:35 GMT+02:00 J. Bruce Fields <[email protected]>:
>> > On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
>> >> Open issues in nfs:
>> >>
>> >> * When a user or group name cannot be mapped, nfs's idmapper always maps it
>> >> to nobody. That's good enough for mapping the file owner and owning
>> >> group, but not for identifiers in acls. For now, to get the nfs richacl
>> >> support somewhat working, I'm explicitly checking if mapping has resulted
>> >> in uid/gid 99 in the kernel.
>> >>
>> >> * When the nfs server replies with NFS4ERR_BADNAME for any user or group
>> >> name lookup, the client will stop sending numeric uids and gids to the
>> >> server even when the lookup wasn't numeric. From then on, the client
>> >> will translate uids and gids that have no mapping to the string "nobody",
>> >> and the server will reject them. This problem is not specific to acls.
>> >
>> > Do you have fixes in mind for these two issues?
>>
>> I'm not sure how to best fix the idmapper problem, with backwards
>> compatibility and all.
>
> I haven't looked at the current nfsidmap interface.... So it's
> completely lacking any way to communicate failure?

Yes, when a user doesn't exist, idmapper maps that to the nobody
uid/gid. That's the failure mode of stat. In the acl case, we do want
to map user and group names to their respective ids where possible (so
that the acl makes sense in the local system context), but we do want
to preserve the original user and group names when there is no such
mapping instead of mapping to the nobody uid/gid.

>> The second problem shouldn't be too hard to fix.
>
> Is it enough to turn off the failover in the case there's no possibility
> it could have been caused by a numeric id?

Yes, I believe that would be enough.

> If any user can set ACLs with arbitrary strings as names, then we'd be
> giving any user unprivileged user the ability to turn off numeric
> idmapping, so I think we need to fix that.

The bug can be triggered by unprivileged users with nfs4_setfacl.

Thanks,
Andreas

2015-10-04 06:23:32

by Christoph Hellwig

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
> Hello,
>
> here's another update of the richacl patch queue. At this stage, I would
> like to ask for final feedback so that the core and ext4 code (patches
> 1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
> then go through the respective maintainer trees.

Now way in this form even if everyone agrees we should have these
bastard ACLs. I certainly disagree.

Ayway, back to the VFS <-> FS interface. You still require tons of
boilderplate code in the filesystem which isn't required and we got rid
of for Posix ACLs. The filesystem should not look at the userspace
xattr format, please follow a model similar to ->get_acl and ->set_acl
for Posix ACLs. After that the wire up should be so trivial that you
can wire up btrfs, xfs and f2fs as well, which is important to make the
feature mergeable.

And honestly I tink adding even more overload to xattrs is a really bad
idea and after 10 years of experience with that junk we really need to
learn and make new overloads proper system calls.


2015-10-05 18:45:42

by Andreas Gruenbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Sun, Oct 4, 2015 at 8:23 AM, Christoph Hellwig <[email protected]> wrote:
> On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
>> Hello,
>>
>> here's another update of the richacl patch queue. At this stage, I would
>> like to ask for final feedback so that the core and ext4 code (patches
>> 1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
>> then go through the respective maintainer trees.
>
> Now way in this form even if everyone agrees we should have these
> bastard ACLs. I certainly disagree.

Well, thanks for having a look at the patches.

> Ayway, back to the VFS <-> FS interface. You still require tons of
> boilderplate code in the filesystem which isn't required and we got rid
> of for Posix ACLs. The filesystem should not look at the userspace
> xattr format, please follow a model similar to ->get_acl and ->set_acl
> for Posix ACLs.

I will repost a version that has this cleaned up.

> After that the wire up should be so trivial that you can wire up btrfs,
> xfs and f2fs as well, which is important to make the feature mergeable.

Why would the patch queue become more mergeable by having support for
more filesystems in it? The filesystem specific code really isn't all
that interesting.

> And honestly I tink adding even more overload to xattrs is a really bad
> idea and after 10 years of experience with that junk we really need to
> learn and make new overloads proper system calls.

Thanks,
Andreas

2015-10-05 18:59:21

by Austin S Hemmelgarn

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On 2015-10-05 14:45, Andreas Gruenbacher wrote:
> On Sun, Oct 4, 2015 at 8:23 AM, Christoph Hellwig <[email protected]> wrote:
>> On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
>>> Hello,
>>>
>>> here's another update of the richacl patch queue. At this stage, I would
>>> like to ask for final feedback so that the core and ext4 code (patches
>>> 1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
>>> then go through the respective maintainer trees.
>>
>> Now way in this form even if everyone agrees we should have these
>> bastard ACLs. I certainly disagree.
>
> Well, thanks for having a look at the patches.
>
>> Ayway, back to the VFS <-> FS interface. You still require tons of
>> boilderplate code in the filesystem which isn't required and we got rid
>> of for Posix ACLs. The filesystem should not look at the userspace
>> xattr format, please follow a model similar to ->get_acl and ->set_acl
>> for Posix ACLs.
>
> I will repost a version that has this cleaned up.
>
>> After that the wire up should be so trivial that you can wire up btrfs,
>> xfs and f2fs as well, which is important to make the feature mergeable.
>
> Why would the patch queue become more mergeable by having support for
> more filesystems in it? The filesystem specific code really isn't all
> that interesting.
I think the point is that a new VFS feature that is easy to integrate in
multiple filesystems should have support for those filesystems. A
decade ago, just having ext* support would probably have been fine, but
these days, XFS, BTRFS, and F2FS are used just as much (if not more) on
production systems as ext4, and having support for them right from the
start would significantly help with adoption of richacls.
>
>> And honestly I tink adding even more overload to xattrs is a really bad
>> idea and after 10 years of experience with that junk we really need to
>> learn and make new overloads proper system calls.
>
> Thanks,
> Andreas
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>



Attachments:
smime.p7s (2.95 kB)
S/MIME Cryptographic Signature

2015-10-05 21:17:15

by Dave Chinner

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Mon, Oct 05, 2015 at 08:45:40PM +0200, Andreas Gruenbacher wrote:
> On Sun, Oct 4, 2015 at 8:23 AM, Christoph Hellwig <[email protected]> wrote:
> > On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
> >> Hello,
> >>
> >> here's another update of the richacl patch queue. At this stage, I would
> >> like to ask for final feedback so that the core and ext4 code (patches
> >> 1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
> >> then go through the respective maintainer trees.
> >
> > Now way in this form even if everyone agrees we should have these
> > bastard ACLs. I certainly disagree.
>
> Well, thanks for having a look at the patches.
>
> > Ayway, back to the VFS <-> FS interface. You still require tons of
> > boilderplate code in the filesystem which isn't required and we got rid
> > of for Posix ACLs. The filesystem should not look at the userspace
> > xattr format, please follow a model similar to ->get_acl and ->set_acl
> > for Posix ACLs.
>
> I will repost a version that has this cleaned up.
>
> > After that the wire up should be so trivial that you can wire up btrfs,
> > xfs and f2fs as well, which is important to make the feature mergeable.
>
> Why would the patch queue become more mergeable by having support for
> more filesystems in it? The filesystem specific code really isn't all
> that interesting.

The hardest part for the filesystem support is the on-disk feature
flag that needs to be set. The kernel part of that is easy, but it's
an on-disk format change and so there's also all the userspace side
for mkfs, fsck, debug tools, etc, that also need to be able to parse
and understand it. So while the xattr code can be made much more
generic, there's a bunch of filesystem specific code that needs to
go into multiple different repositories and userspace packages for
this.

Andreas, I also can't remember if any xfstests have been written for
these ACLs? That would certainly help make sure all these
filesystems have equivalent behaviour...

Cheers,

Dave.
--
Dave Chinner
[email protected]

2015-10-05 22:01:21

by Andreas Gruenbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Mon, Oct 5, 2015 at 11:17 PM, Dave Chinner <[email protected]> wrote:
> On Mon, Oct 05, 2015 at 08:45:40PM +0200, Andreas Gruenbacher wrote:
>> On Sun, Oct 4, 2015 at 8:23 AM, Christoph Hellwig <[email protected]> wrote:
>> > After that the wire up should be so trivial that you can wire up btrfs,
>> > xfs and f2fs as well, which is important to make the feature mergeable.
>>
>> Why would the patch queue become more mergeable by having support for
>> more filesystems in it? The filesystem specific code really isn't all
>> that interesting.
>
> The hardest part for the filesystem support is the on-disk feature
> flag that needs to be set. The kernel part of that is easy, but it's
> an on-disk format change and so there's also all the userspace side
> for mkfs, fsck, debug tools, etc, that also need to be able to parse
> and understand it. So while the xattr code can be made much more
> generic, there's a bunch of filesystem specific code that needs to
> go into multiple different repositories and userspace packages for
> this.

Yes.

> Andreas, I also can't remember if any xfstests have been written for
> these ACLs? That would certainly help make sure all these
> filesystems have equivalent behaviour...

There's a reasonable amount of tests in the richacl user-space package
which are shell based, with a few small C helpers. We could move those
into xfstests eventually; now seems a bit early to me.

Thanks,
Andreas

2015-10-06 01:20:25

by Dave Chinner

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Tue, Oct 06, 2015 at 12:01:19AM +0200, Andreas Gruenbacher wrote:
> On Mon, Oct 5, 2015 at 11:17 PM, Dave Chinner <[email protected]> wrote:
> > On Mon, Oct 05, 2015 at 08:45:40PM +0200, Andreas Gruenbacher wrote:
> >> On Sun, Oct 4, 2015 at 8:23 AM, Christoph Hellwig <[email protected]> wrote:
> >> > After that the wire up should be so trivial that you can wire up btrfs,
> >> > xfs and f2fs as well, which is important to make the feature mergeable.
> >>
> >> Why would the patch queue become more mergeable by having support for
> >> more filesystems in it? The filesystem specific code really isn't all
> >> that interesting.
> >
> > The hardest part for the filesystem support is the on-disk feature
> > flag that needs to be set. The kernel part of that is easy, but it's
> > an on-disk format change and so there's also all the userspace side
> > for mkfs, fsck, debug tools, etc, that also need to be able to parse
> > and understand it. So while the xattr code can be made much more
> > generic, there's a bunch of filesystem specific code that needs to
> > go into multiple different repositories and userspace packages for
> > this.
>
> Yes.
>
> > Andreas, I also can't remember if any xfstests have been written for
> > these ACLs? That would certainly help make sure all these
> > filesystems have equivalent behaviour...
>
> There's a reasonable amount of tests in the richacl user-space package
> which are shell based, with a few small C helpers. We could move those
> into xfstests eventually; now seems a bit early to me.

Well, all the fs developers that will do the userspace work are
already running xfstests. If you want us to be able to test the
richACL code as we add all the fs specific flags to the userspace
code, then we need the tests in xfstests at the same time the
infrastructure appears in the kernel...

Cheers,

Dave.
--
Dave Chinner
[email protected]

2015-10-06 09:49:41

by James Morris

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Sat, 3 Oct 2015, Christoph Hellwig wrote:

> On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
> > Hello,
> >
> > here's another update of the richacl patch queue. At this stage, I would
> > like to ask for final feedback so that the core and ext4 code (patches
> > 1-19) can be merged in the 4.4 merge window. The nfsd and nfs code should
> > then go through the respective maintainer trees.
>
> Now way in this form even if everyone agrees we should have these
> bastard ACLs. I certainly disagree.
>

Where is the rationale for them?

This url doesn't work: http://acl.bestbits.at/richacl/





--
James Morris
<[email protected]>


2015-10-06 09:49:59

by Christoph Hellwig

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Mon, Oct 05, 2015 at 02:58:36PM -0400, Austin S Hemmelgarn wrote:
> I think the point is that a new VFS feature that is easy to integrate in
> multiple filesystems should have support for those filesystems. A decade
> ago, just having ext* support would probably have been fine, but these days,
> XFS, BTRFS, and F2FS are used just as much (if not more) on production
> systems as ext4, and having support for them right from the start would
> significantly help with adoption of richacls.

That's one reason. The other is that actually wiring it up for more
than a single consumer shows its actually reasonable generic. I don't
want to end up with a situration like Posix ACLs again where different
file systems using different on disk formats again.

2015-10-06 12:24:11

by Andreas Grünbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

2015-10-06 11:49 GMT+02:00 James Morris <[email protected]>:
> Where is the rationale for them?
>
> This url doesn't work: http://acl.bestbits.at/richacl/

That URL also properly redirects to http://www.bestbits.at/richacl/
now. There is some information on that website, in the manual pages in
the richacl user-space package, an in the patch descriptions and
comments.

Thanks,
Andreas

2015-10-06 13:12:38

by Andreas Gruenbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Tue, Oct 6, 2015 at 11:49 AM, Christoph Hellwig <[email protected]> wrote:
> On Mon, Oct 05, 2015 at 02:58:36PM -0400, Austin S Hemmelgarn wrote:
>> I think the point is that a new VFS feature that is easy to integrate in
>> multiple filesystems should have support for those filesystems. A decade
>> ago, just having ext* support would probably have been fine, but these days,
>> XFS, BTRFS, and F2FS are used just as much (if not more) on production
>> systems as ext4, and having support for them right from the start would
>> significantly help with adoption of richacls.
>
> That's one reason. The other is that actually wiring it up for more
> than a single consumer shows its actually reasonable generic.

The filesystem interface now is the same as for POSIX ACLs, used by a
dozen or so filesystems already.

> I don't want to end up with a situration like Posix ACLs again where
> different file systems using different on disk formats again.

Any file system could choose a different on-disk format than the one
that ext4 currently uses, but I don't see a reason why any should.
Apart from uid / gid mappings that is the same as the user-space xattr
format. Network file systems like NFSv4 and CIFS with their predefined
over-the-wire formats obviously are another story.

Thanks,
Andreas

2015-10-06 20:26:18

by Andreas Dilger

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Oct 6, 2015, at 7:12 AM, Andreas Gruenbacher <[email protected]> wrote:
>
> On Tue, Oct 6, 2015 at 11:49 AM, Christoph Hellwig <[email protected]> wrote:
>> On Mon, Oct 05, 2015 at 02:58:36PM -0400, Austin S Hemmelgarn wrote:
>>> I think the point is that a new VFS feature that is easy to integrate in
>>> multiple filesystems should have support for those filesystems. A decade
>>> ago, just having ext* support would probably have been fine, but these days,
>>> XFS, BTRFS, and F2FS are used just as much (if not more) on production
>>> systems as ext4, and having support for them right from the start would
>>> significantly help with adoption of richacls.
>>
>> That's one reason. The other is that actually wiring it up for more
>> than a single consumer shows its actually reasonable generic.
>
> The filesystem interface now is the same as for POSIX ACLs, used by a
> dozen or so filesystems already.
>
>> I don't want to end up with a situration like Posix ACLs again where
>> different file systems using different on disk formats again.
>
> Any file system could choose a different on-disk format than the one
> that ext4 currently uses, but I don't see a reason why any should.
> Apart from uid / gid mappings that is the same as the user-space xattr
> format. Network file systems like NFSv4 and CIFS with their predefined
> over-the-wire formats obviously are another story.

And any disk filesystems that have their own non-POSIX ACLs, such as HFS, NTFS, ZFS would presumably also need to map the in-kernel Richacl format to their on-disk format.

Cheers, Andreas






Attachments:
signature.asc (833.00 B)
Message signed with OpenPGP using GPGMail

2015-10-06 20:55:21

by Steve French

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Tue, Oct 6, 2015 at 3:26 PM, Andreas Dilger <[email protected]> wrote:
> On Oct 6, 2015, at 7:12 AM, Andreas Gruenbacher <[email protected]> wrote:
>>
>> On Tue, Oct 6, 2015 at 11:49 AM, Christoph Hellwig <[email protected]> wrote:
>>> On Mon, Oct 05, 2015 at 02:58:36PM -0400, Austin S Hemmelgarn wrote:
>>>> I think the point is that a new VFS feature that is easy to integrate in
>>>> multiple filesystems should have support for those filesystems. A decade
>>>> ago, just having ext* support would probably have been fine, but these days,
>>>> XFS, BTRFS, and F2FS are used just as much (if not more) on production
>>>> systems as ext4, and having support for them right from the start would
>>>> significantly help with adoption of richacls.
>>>
>>> That's one reason. The other is that actually wiring it up for more
>>> than a single consumer shows its actually reasonable generic.
>>
>> The filesystem interface now is the same as for POSIX ACLs, used by a
>> dozen or so filesystems already.
>>
>>> I don't want to end up with a situration like Posix ACLs again where
>>> different file systems using different on disk formats again.
>>
>> Any file system could choose a different on-disk format than the one
>> that ext4 currently uses, but I don't see a reason why any should.
>> Apart from uid / gid mappings that is the same as the user-space xattr
>> format. Network file systems like NFSv4 and CIFS with their predefined
>> over-the-wire formats obviously are another story.
>
> And any disk filesystems that have their own non-POSIX ACLs, such as HFS, NTFS, ZFS would presumably also need to map the in-kernel Richacl format to their on-disk format.

Will be interesting to see whether in the long run can have some
common code in NTFS, CIFS/SMB3 (and even NFSv4.x) ACL parsing since
their formats are quite closely related



--
Thanks,

Steve

2015-10-07 07:51:08

by Christoph Hellwig

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Tue, Oct 06, 2015 at 02:26:09PM -0600, Andreas Dilger wrote:
> And any disk filesystems that have their own non-POSIX ACLs, such as HFS, NTFS, ZFS would presumably also need to map the in-kernel Richacl format to their on-disk format.

No, we did this mistake with Posix ACLs, and we're not going to repeat
it here. Filesystems with their own slightly different ACLs must not
reuse the interface.

2015-10-07 13:38:50

by Andreas Gruenbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Wed, Oct 7, 2015 at 9:50 AM, Christoph Hellwig <[email protected]> wrote:
> On Tue, Oct 06, 2015 at 02:26:09PM -0600, Andreas Dilger wrote:
>> And any disk filesystems that have their own non-POSIX ACLs, such as HFS, NTFS, ZFS would presumably also need to map the in-kernel Richacl format to their on-disk format.
>
> No, we did this mistake with Posix ACLs, and we're not going to repeat
> it here. Filesystems with their own slightly different ACLs must not
> reuse the interface.

Well, things may not be quite as clearly delineated. We currently have
code in nfsd for mapping between NFSv4 ACLs on the wire and POSIX ACLs
on local file systems. This mapping is problematic because of the
semantic differences between NFSv4 ACLs and POSIX ACLs (different sets
of permissions, access and default acl vs. inheritance flags,
different permission check algorithm). I wish we could have avoided
that.

Richacls are designed to support NFSv4 ACLs on top of POSIX systems.
This means that they should obviously be supported by the NFSv4 server
and client (see the patches) and by the common local filesystems.

ACLs on NTFS and ZFS mostly fit into the same model. The big remaining
difference there is how users and groups are identified: NTFS used
SIDs (https://en.wikipedia.org/wiki/Security_Identifier); ZFS could be
said to use a hybrid UID / GID / SID model. Exposing those ACLs as
richacls would make sense if we can find a clean way of handling this
aspect.

HFS ACLs have sufficiently different semantics (the user.group tuples)
that representing them as richacls wouldn't make sense.

Thanks,
Andreas

2015-10-07 15:15:40

by Steve French

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Wed, Oct 7, 2015 at 8:38 AM, Andreas Gruenbacher <[email protected]> wrote:
> On Wed, Oct 7, 2015 at 9:50 AM, Christoph Hellwig <[email protected]> wrote:
>> On Tue, Oct 06, 2015 at 02:26:09PM -0600, Andreas Dilger wrote:
>>> And any disk filesystems that have their own non-POSIX ACLs, such as HFS, NTFS, ZFS would presumably also need to map the in-kernel Richacl format to their on-disk format.
>>
>> No, we did this mistake with Posix ACLs, and we're not going to repeat
>> it here. Filesystems with their own slightly different ACLs must not
>> reuse the interface.
>
> Well, things may not be quite as clearly delineated. We currently have
> code in nfsd for mapping between NFSv4 ACLs on the wire and POSIX ACLs
> on local file systems. This mapping is problematic because of the
> semantic differences between NFSv4 ACLs and POSIX ACLs (different sets
> of permissions, access and default acl vs. inheritance flags,
> different permission check algorithm). I wish we could have avoided
> that.
>
> Richacls are designed to support NFSv4 ACLs on top of POSIX systems.
> This means that they should obviously be supported by the NFSv4 server
> and client (see the patches) and by the common local filesystems.
>
> ACLs on NTFS and ZFS mostly fit into the same model. The big remaining
> difference there is how users and groups are identified: NTFS used
> SIDs (https://en.wikipedia.org/wiki/Security_Identifier); ZFS could be
> said to use a hybrid UID / GID / SID model. Exposing those ACLs as
> richacls would make sense if we can find a clean way of handling this
> aspect.

Samba (e.g. winbind service) has mapping libraries for mapping SIDs to
UIDs (CIFS ACLs already have the same issue of SID to UID mapping
which we handle with upcalls) and Samba has various pluggable ways to
handle UID mapping and is easily extensible. Similarly NFSv4 ACLs,
although closely related to CIFS/NTFS ACLs have to be map usernames to
uids.

--
Thanks,

Steve

2015-10-16 18:12:40

by Andreas Gruenbacher

[permalink] [raw]
Subject: Re: [PATCH v8 00/41] Richacls

On Tue, Sep 29, 2015 at 4:54 PM, Andreas Grünbacher
<[email protected]> wrote:
> 2015-09-28 19:46 GMT+02:00 J. Bruce Fields <[email protected]>:
>> On Mon, Sep 28, 2015 at 07:10:06PM +0200, Andreas Grünbacher wrote:
>>> 2015-09-28 18:35 GMT+02:00 J. Bruce Fields <[email protected]>:
>>> > On Mon, Sep 28, 2015 at 12:08:51AM +0200, Andreas Gruenbacher wrote:
>>> >> Open issues in nfs:
>>> >>
>>> >> * When a user or group name cannot be mapped, nfs's idmapper always maps it
>>> >> to nobody. That's good enough for mapping the file owner and owning
>>> >> group, but not for identifiers in acls. For now, to get the nfs richacl
>>> >> support somewhat working, I'm explicitly checking if mapping has resulted
>>> >> in uid/gid 99 in the kernel.
>>> >>
>>> >> * When the nfs server replies with NFS4ERR_BADNAME for any user or group
>>> >> name lookup, the client will stop sending numeric uids and gids to the
>>> >> server even when the lookup wasn't numeric. From then on, the client
>>> >> will translate uids and gids that have no mapping to the string "nobody",
>>> >> and the server will reject them. This problem is not specific to acls.
>>> >
>>> > Do you have fixes in mind for these two issues?
>>>
>>> I'm not sure how to best fix the idmapper problem, with backwards
>>> compatibility and all.
>>
>> I haven't looked at the current nfsidmap interface.... So it's
>> completely lacking any way to communicate failure?
>
> Yes, when a user doesn't exist, idmapper maps that to the nobody
> uid/gid. That's the failure mode of stat. In the acl case, we do want
> to map user and group names to their respective ids where possible (so
> that the acl makes sense in the local system context), but we do want
> to preserve the original user and group names when there is no such
> mapping instead of mapping to the nobody uid/gid.

So that's fixed now in nfs-utils and the latest richacl snapshot.

>>> The second problem shouldn't be too hard to fix.
>>
>> Is it enough to turn off the failover in the case there's no possibility
>> it could have been caused by a numeric id?
>
> Yes, I believe that would be enough.
>
>> If any user can set ACLs with arbitrary strings as names, then we'd be
>> giving any user unprivileged user the ability to turn off numeric
>> idmapping, so I think we need to fix that.
>
> The bug can be triggered by unprivileged users with nfs4_setfacl.

It turns out that when setting an ACL, the ACL can contain UID/GID
numbers as well as user@domain strings which the server cannot map.
The status code is NFS4ERR_BADNAME both when the server doesn't
support UID/GID numbers and when it cannot map a name, so in that
case, we cannot tell the difference.

Luckily, when a UID/GID cannot be mapped to a name, nfs falls back to
sending the server a UID/GID number instead even when
NFS_CAP_UIDGID_NOMAP has been cleared. So unprivileged users can turn
on the idmapper, but at least that doesn't break unmapped UIDs/GIDs.
(I got that wrong in my initial analysis.)

Thanks,
Andreas