linux/fs
Michael Halcrow 88b4a07e66 [PATCH] eCryptfs: Public key transport mechanism
This is the transport code for public key functionality in eCryptfs.  It
manages encryption/decryption request queues with a transport mechanism.
Currently, netlink is the only implemented transport.

Each inode has a unique File Encryption Key (FEK).  Under passphrase, a File
Encryption Key Encryption Key (FEKEK) is generated from a salt/passphrase
combo on mount.  This FEKEK encrypts each FEK and writes it into the header of
each file using the packet format specified in RFC 2440.  This is all
symmetric key encryption, so it can all be done via the kernel crypto API.

These new patches introduce public key encryption of the FEK.  There is no
asymmetric key encryption support in the kernel crypto API, so eCryptfs pushes
the FEK encryption and decryption out to a userspace daemon.  After
considering our requirements and determining the complexity of using various
transport mechanisms, we settled on netlink for this communication.

eCryptfs stores authentication tokens into the kernel keyring.  These tokens
correlate with individual keys.  For passphrase mode of operation, the
authentication token contains the symmetric FEKEK.  For public key, the
authentication token contains a PKI type and an opaque data blob managed by
individual PKI modules in userspace.

Each user who opens a file under an eCryptfs partition mounted in public key
mode must be running a daemon.  That daemon has the user's credentials and has
access to all of the keys to which the user should have access.  The daemon,
when started, initializes the pluggable PKI modules available on the system
and registers itself with the eCryptfs kernel module.  Userspace utilities
register public key authentication tokens into the user session keyring.
These authentication tokens correlate key signatures with PKI modules and PKI
blobs.  The PKI blobs contain PKI-specific information necessary for the PKI
module to carry out asymmetric key encryption and decryption.

When the eCryptfs module parses the header of an existing file and finds a Tag
1 (Public Key) packet (see RFC 2440), it reads in the public key identifier
(signature).  The asymmetrically encrypted FEK is in the Tag 1 packet;
eCryptfs puts together a decrypt request packet containing the signature and
the encrypted FEK, then it passes it to the daemon registered for the
current->euid via a netlink unicast to the PID of the daemon, which was
registered at the time the daemon was started by the user.

The daemon actually just makes calls to libecryptfs, which implements request
packet parsing and manages PKI modules.  libecryptfs grabs the public key
authentication token for the given signature from the user session keyring.
This auth tok tells libecryptfs which PKI module should receive the request.
libecryptfs then makes a decrypt() call to the PKI module, and it passes along
the PKI block from the auth tok.  The PKI uses the blob to figure out how it
should decrypt the data passed to it; it performs the decryption and passes
the decrypted data back to libecryptfs.  libecryptfs then puts together a
reply packet with the decrypted FEK and passes that back to the eCryptfs
module.

The eCryptfs module manages these request callouts to userspace code via
message context structs.  The module maintains an array of message context
structs and places the elements of the array on two lists: a free and an
allocated list.  When eCryptfs wants to make a request, it moves a msg ctx
from the free list to the allocated list, sets its state to pending, and fires
off the message to the user's registered daemon.

When eCryptfs receives a netlink message (via the callback), it correlates the
msg ctx struct in the alloc list with the data in the message itself.  The
msg->index contains the offset of the array of msg ctx structs.  It verifies
that the registered daemon PID is the same as the PID of the process that sent
the message.  It also validates a sequence number between the received packet
and the msg ctx.  Then, it copies the contents of the message (the reply
packet) into the msg ctx struct, sets the state in the msg ctx to done, and
wakes up the process that was sleeping while waiting for the reply.

The sleeping process was whatever was performing the sys_open().  This process
originally called ecryptfs_send_message(); it is now in
ecryptfs_wait_for_response().  When it wakes up and sees that the msg ctx
state was set to done, it returns a pointer to the message contents (the reply
packet) and returns.  If all went well, this packet contains the decrypted
FEK, which is then copied into the crypt_stat struct, and life continues as
normal.

The case for creation of a new file is very similar, only instead of a decrypt
request, eCryptfs sends out an encrypt request.

> - We have a great clod of key mangement code in-kernel.  Why is that
>   not suitable (or growable) for public key management?

eCryptfs uses Howells' keyring to store persistent key data and PKI state
information.  It defers public key cryptographic transformations to userspace
code.  The userspace data manipulation request really is orthogonal to key
management in and of itself.  What eCryptfs basically needs is a secure way to
communicate with a particular daemon for a particular task doing a syscall,
based on the UID.  Nothing running under another UID should be able to access
that channel of communication.

> - Is it appropriate that new infrastructure for public key
> management be private to a particular fs?

The messaging.c file contains a lot of code that, perhaps, could be extracted
into a separate kernel service.  In essence, this would be a sort of
request/reply mechanism that would involve a userspace daemon.  I am not aware
of anything that does quite what eCryptfs does, so I was not aware of any
existing tools to do just what we wanted.

>   What happens if one of these daemons exits without sending a quit
>   message?

There is a stale uid<->pid association in the hash table for that user.  When
the user registers a new daemon, eCryptfs cleans up the old association and
generates a new one.  See ecryptfs_process_helo().

> - _why_ does it use netlink?

Netlink provides the transport mechanism that would minimize the complexity of
the implementation, given that we can have multiple daemons (one per user).  I
explored the possibility of using relayfs, but that would involve having to
introduce control channels and a protocol for creating and tearing down
channels for the daemons.  We do not have to worry about any of that with
netlink.

Signed-off-by: Michael Halcrow <mhalcrow@us.ibm.com>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-02-12 09:48:36 -08:00
..
9p [PATCH] remove invalidate_inode_pages() 2007-02-11 10:51:31 -08:00
adfs [PATCH] adfs: fix filename handling 2007-01-05 23:55:22 -08:00
affs [PATCH] affs: change uses of f_{dentry, vfsmnt} to use f_path 2006-12-08 08:28:43 -08:00
afs [PATCH] rename struct namespace to struct mnt_namespace 2006-12-08 08:28:51 -08:00
autofs [PATCH] autofs: change uses of f_{dentry, vfsmnt} to use f_path 2006-12-08 08:28:43 -08:00
autofs4 [PATCH] getting rid of all casts of k[cmz]alloc() calls 2006-12-13 09:05:58 -08:00
befs [PATCH] getting rid of all casts of k[cmz]alloc() calls 2006-12-13 09:05:58 -08:00
bfs [PATCH] update Tigran's email addresses 2006-12-13 09:05:53 -08:00
cifs [CIFS] Minor cleanup 2007-02-06 20:43:30 +00:00
coda [PATCH] seq_file conversion: coda 2007-02-11 10:51:34 -08:00
configfs [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
cramfs [PATCH] struct path: convert cramfs 2006-12-08 08:28:44 -08:00
debugfs DebugFS : file/directory removal fix 2006-12-13 15:38:45 -08:00
devpts [PATCH] inode-diet: Eliminate i_blksize from the inode structure 2006-09-27 08:26:18 -07:00
dlm [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
ecryptfs [PATCH] eCryptfs: Public key transport mechanism 2007-02-12 09:48:36 -08:00
efs [PATCH] struct path: convert efs 2006-12-08 08:28:45 -08:00
exportfs [PATCH] VFS: Make filldir_t and struct kstat deal in 64-bit inode numbers 2006-10-03 08:03:40 -07:00
ext2 [PATCH] fix umask when noACL kernel meets extN tuned for ACLs 2007-02-11 10:51:34 -08:00
ext3 [PATCH] jbd layer function called instead of fs specific one 2007-02-11 11:18:06 -08:00
ext4 [PATCH] jbd layer function called instead of fs specific one 2007-02-11 11:18:06 -08:00
fat [PATCH] fat: change uses of f_{dentry,vfsmnt} to use f_path 2006-12-08 08:28:41 -08:00
freevxfs [PATCH] struct path: convert freevxfs 2006-12-08 08:28:45 -08:00
fuse [PATCH] remove invalidate_inode_pages() 2007-02-11 10:51:31 -08:00
gfs2 [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
hfs [PATCH] extend the set of "__attribute__" shortcut macros 2007-02-11 10:51:35 -08:00
hfsplus [PATCH] extend the set of "__attribute__" shortcut macros 2007-02-11 10:51:35 -08:00
hostfs [PATCH] uml: fix mknod 2007-01-30 08:26:44 -08:00
hpfs [PATCH] struct path: convert hpfs 2006-12-08 08:28:45 -08:00
hppfs [PATCH] struct path: convert hppfs 2006-12-08 08:28:45 -08:00
hugetlbfs [PATCH] hugetlb: preserve hugetlb pte dirty state 2007-02-09 09:25:46 -08:00
isofs [PATCH] isofs: change uses of f_{dentry, vfsmnt} to use f_path 2006-12-08 08:28:41 -08:00
jbd [PATCH] jbd: wait for already submitted t_sync_datalist buffer to complete 2006-12-22 08:55:51 -08:00
jbd2 [PATCH] jbd2: wait for already submitted t_sync_datalist buffer to complete 2006-12-07 08:39:42 -08:00
jffs [PATCH] remove invalidate_inode_pages() 2007-02-11 10:51:31 -08:00
jffs2 Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 2007-01-18 10:34:51 +11:00
jfs JFS: Remove incorrect kgdb define 2007-01-26 10:14:36 -06:00
lockd [PATCH] knfsd: SUNRPC: Provide room in svc_rqst for larger addresses 2007-02-12 09:48:36 -08:00
minix [PATCH] Minix V3 support 2007-02-12 09:48:31 -08:00
msdos [PATCH] fat: add fat_getattr() 2006-11-16 11:43:38 -08:00
ncpfs [PATCH] ncpfs: ensure we free wdog_pid on parse_option or fill_inode failure 2006-12-13 09:05:53 -08:00
nfs [PATCH] include/linux/nfsd/const.h: remove NFS_SUPER_MAGIC 2007-02-12 09:48:36 -08:00
nfs_common [PATCH] nfs_common endianness annotations 2006-10-20 10:26:41 -07:00
nfsd [PATCH] knfsd: SUNRPC: Provide room in svc_rqst for larger addresses 2007-02-12 09:48:36 -08:00
nls [PATCH] fs: make nls_cp936.c handle some U00XY characters and U20AC correctly 2006-12-07 08:39:46 -08:00
ntfs [PATCH] NTFS: rename incorrect check of NTFS_DEBUG with just DEBUG 2007-02-12 09:48:27 -08:00
ocfs2 [PATCH] ocfs2 heartbeat: clean up bio submission code 2007-02-07 12:15:58 -08:00
openpromfs [PATCH] struct path: convert openpromfs 2006-12-08 08:28:48 -08:00
partitions Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6 2007-02-11 11:37:45 -08:00
proc [PATCH] tty: update the tty layer to work with struct pid 2007-02-12 09:48:32 -08:00
qnx4 [PATCH] struct path: convert qnx4 2006-12-08 08:28:48 -08:00
ramfs [PATCH] convert ramfs to use __set_page_dirty_no_writeback 2007-02-11 10:51:19 -08:00
reiserfs [PATCH] reiserfs: Use ARRAY_SIZE macro when appropriate 2007-02-12 09:48:29 -08:00
romfs [PATCH] struct path: convert romfs 2006-12-08 08:28:49 -08:00
smbfs [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
sysfs [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
sysv [PATCH] fs/sysv/: proper prototypes for 2 functions 2006-12-22 08:55:47 -08:00
udf [PATCH] struct path: convert udf 2006-12-08 08:28:50 -08:00
ufs [PATCH] ufs: restore back support of openstep 2007-02-09 09:25:46 -08:00
vfat [PATCH] fat: add fat_getattr() 2006-11-16 11:43:38 -08:00
xfs [PATCH] Make XFS use BH_Unwritten and BH_Delay correctly 2007-02-12 09:48:27 -08:00
Kconfig [PATCH] eCryptfs: Public key transport mechanism 2007-02-12 09:48:36 -08:00
Kconfig.binfmt
Makefile [PATCH] fsstack: Introduce fsstack_copy_{attr,inode}_* 2006-12-08 08:28:40 -08:00
aio.c [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
attr.c
bad_inode.c [PATCH] fix memory corruption from misinterpreted bad_inode_ops return values 2007-01-05 23:55:23 -08:00
binfmt_aout.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
binfmt_elf.c [PATCH] core-dumping unreadable binaries via PT_INTERP 2007-01-26 13:51:00 -08:00
binfmt_elf_fdpic.c [PATCH] Remove final references to deprecated "MAP_ANON" page protection flag 2007-02-11 10:51:17 -08:00
binfmt_em86.c
binfmt_flat.c [PATCH] uclinux: correctly remap bin_fmtflat exe allocated mem regions 2007-02-09 10:45:33 -08:00
binfmt_misc.c [PATCH] getting rid of all casts of k[cmz]alloc() calls 2006-12-13 09:05:58 -08:00
binfmt_script.c
binfmt_som.c [PARISC] Fix fs/binfmt_som.c 2006-10-04 06:51:26 -06:00
bio.c [PATCH] optimize o_direct on block devices 2006-12-13 09:05:50 -08:00
block_dev.c [PATCH] revert blockdev direct io back to 2.6.19 version 2007-02-03 11:26:06 -08:00
buffer.c [PATCH] warning fix: unsigned->signed 2007-02-12 09:48:29 -08:00
char_dev.c [PATCH] register_chrdev_region() don't hand out the LOCAL/EXPERIMENTAL majors 2007-02-12 09:48:27 -08:00
compat.c [PATCH] fdtable: Make fdarray and fdsets equal in size 2006-12-10 09:57:22 -08:00
compat_ioctl.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
dcache.c [PATCH] Fix d_path for lazy unmounts 2007-02-12 09:48:27 -08:00
dcookies.c [PATCH] slab: remove kmem_cache_t 2006-12-07 08:39:25 -08:00
direct-io.c [PATCH] dio: lock refcount operations 2006-12-10 09:57:21 -08:00
dnotify.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
dquot.c [PATCH] remove sb->s_files and file_list_lock usage in dquot.c 2007-02-12 09:48:28 -08:00
drop_caches.c [PATCH] remove invalidate_inode_pages() 2007-02-11 10:51:31 -08:00
eventpoll.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
exec.c [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
fcntl.c [PATCH] fdtable: Make fdarray and fdsets equal in size 2006-12-10 09:57:22 -08:00
fifo.c [PATCH] pipe.c/fifo.c code cleanups 2006-04-11 13:53:33 +02:00
file.c [PATCH] fdtable: Provide free_fdtable() wrapper 2006-12-22 08:55:50 -08:00
file_table.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
filesystems.c [PATCH] Ban register_filesystem(NULL); 2006-09-29 09:18:20 -07:00
fs-writeback.c Write back inode data pages even when the inode itself is locked 2007-01-26 12:53:20 -08:00
generic_acl.c [PATCH] Generic infrastructure for acls 2006-09-29 09:18:24 -07:00
inode.c [PATCH] move remove_dquot_ref to dqout.c 2007-02-12 09:48:28 -08:00
inotify.c [PATCH] severing fs.h, radix-tree.h -> sched.h 2006-12-04 02:00:24 -05:00
inotify_user.c [PATCH] inotify: read return val fix 2007-02-12 09:48:28 -08:00
internal.h [PATCH] CONFIG_BLOCK internal.h cleanups 2006-09-30 20:52:32 +02:00
ioctl.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
ioprio.c [PATCH] pid: replace do/while_each_task_pid with do/while_each_pid_task 2007-02-12 09:48:32 -08:00
libfs.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
locks.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
mbcache.c [PATCH] slab: remove kmem_cache_t 2006-12-07 08:39:25 -08:00
mpage.c [PATCH] BLOCK: Dissociate generic_writepages() from mpage stuff [try #6] 2006-09-30 20:52:26 +02:00
namei.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
namespace.c [PATCH] Transform kmem_cache_alloc()+memset(0) -> kmem_cache_zalloc(). 2007-02-11 10:51:27 -08:00
nfsctl.c Remove obsolete #include <linux/config.h> 2006-06-30 19:25:36 +02:00
no-block.c [PATCH] BLOCK: Make it possible to disable the block layer [try #6] 2006-09-30 20:52:31 +02:00
open.c [PATCH] fdtable: Make fdarray and fdsets equal in size 2006-12-10 09:57:22 -08:00
pipe.c [PATCH] fix leaks on pipe(2) failure exits 2006-12-21 00:16:03 -08:00
pnode.c [PATCH] rename struct namespace to struct mnt_namespace 2006-12-08 08:28:51 -08:00
pnode.h [PATCH] rename struct namespace to struct mnt_namespace 2006-12-08 08:28:51 -08:00
posix_acl.c [PATCH] kmemdup: some users 2006-10-01 00:39:19 -07:00
quota.c [PATCH] BLOCK: Make it possible to disable the block layer [try #6] 2006-09-30 20:52:31 +02:00
quota_v1.c
quota_v2.c
read_write.c [PATCH] FS: speed up rw_verify_area() 2007-02-12 09:48:29 -08:00
read_write.h [PATCH] Remove readv/writev methods and use aio_read/aio_write instead 2006-10-01 00:39:28 -07:00
readdir.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
select.c [PATCH] fdtable: Make fdarray and fdsets equal in size 2006-12-10 09:57:22 -08:00
seq_file.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
splice.c [PATCH] constify pipe_buf_operations 2006-12-13 09:05:47 -08:00
stack.c [PATCH] fsstack: Remove inode copy 2006-12-22 08:55:48 -08:00
stat.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
super.c [PATCH] Revert bd_mount_mutex back to a semaphore 2007-01-11 18:18:21 -08:00
sync.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
utimes.c [PATCH] severing fs.h, radix-tree.h -> sched.h 2006-12-04 02:00:24 -05:00
xattr.c [PATCH] VFS: change struct file to use struct path 2006-12-08 08:28:41 -08:00
xattr_acl.c