This patch fixes error handling in sel_make_bools(), where currently we'd
get a memory leak via security_get_bools() and try to kfree() the wrong
pointer if called again.
Signed-off-by: James Morris <jmorris@namei.org>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch fixes a possible NULL dereference in policydb_destroy, where
p->type_attr_map can be NULL if policydb_destroy is called to clean up a
partially loaded policy upon an error during policy load. Please apply.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The following patch updates the way SELinux classifies and handles IP
based protocols.
Currently, IP sockets are classified by SELinux as being either TCP, UDP
or 'Raw', the latter being a default for IP socket that is not TCP or UDP.
The classification code is out of date and uses only the socket type
parameter to socket(2) to determine the class of IP socket. So, any
socket created with SOCK_STREAM will be classified by SELinux as TCP, and
SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and
SOCK_DCCP are currently ignored by SELinux, which classifies them as
generic sockets, which means they don't even get basic IP level checking.
This patch changes the SELinux IP socket classification logic, so that
only an IPPROTO_IP protocol value passed to socket(2) classify the socket
as TCP or UDP. The patch also drops the check for SOCK_RAW and converts
it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET
are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets).
Note that protocol-specific support for SCTP, DCCP etc. is not addressed
here, we're just getting these protocols checked at the IP layer.
This fixes a reported problem where SCTP sockets were being recognized as
generic SELinux sockets yet still being passed in one case to an IP level
check, which then fails for generic sockets.
It will also fix bugs where any SOCK_STREAM socket is classified as TCP or
any SOCK_DGRAM socket is classified as UDP.
This patch also unifies the way IP sockets classes are determined in
selinux_socket_bind(), so we use the already calculated value instead of
trying to recalculate it.
Signed-off-by: James Morris <jmorris@namei.org>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
With the use of RCU in files structure, the look-up of files using fds can now
be lock-free. The lookup is protected by rcu_read_lock()/rcu_read_unlock().
This patch changes the readers to use lock-free lookup.
Signed-off-by: Maneesh Soni <maneesh@in.ibm.com>
Signed-off-by: Ravikiran Thirumalai <kiran_th@gmail.com>
Signed-off-by: Dipankar Sarma <dipankar@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
In order for the RCU to work, the file table array, sets and their sizes must
be updated atomically. Instead of ensuring this through too many memory
barriers, we put the arrays and their sizes in a separate structure. This
patch takes the first step of putting the file table elements in a separate
structure fdtable that is embedded withing files_struct. It also changes all
the users to refer to the file table using files_fdtable() macro. Subsequent
applciation of RCU becomes easier after this.
Signed-off-by: Dipankar Sarma <dipankar@in.ibm.com>
Signed-Off-By: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch removes the inode_post_link and inode_post_rename LSM hooks as
they are unused (and likely useless).
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch removes the inode_post_create/mkdir/mknod/symlink LSM hooks as
they are obsoleted by the new inode_init_security hook that enables atomic
inode security labeling.
If anyone sees any reason to retain these hooks, please speak now. Also,
is anyone using the post_rename/link hooks; if not, those could also be
removed.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch modifies tmpfs to call the inode_init_security LSM hook to set
up the incore inode security state for new inodes before the inode becomes
accessible via the dcache.
As there is no underlying storage of security xattrs in this case, it is
not necessary for the hook to return the (name, value, len) triple to the
tmpfs code, so this patch also modifies the SELinux hook function to
correctly handle the case where the (name, value, len) pointers are NULL.
The hook call is needed in tmpfs in order to support proper security
labeling of tmpfs inodes (e.g. for udev with tmpfs /dev in Fedora). With
this change in place, we should then be able to remove the
security_inode_post_create/mkdir/... hooks safely.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The following patch set enables atomic security labeling of newly created
inodes by altering the fs code to invoke a new LSM hook to obtain the security
attribute to apply to a newly created inode and to set up the incore inode
security state during the inode creation transaction. This parallels the
existing processing for setting ACLs on newly created inodes. Otherwise, it
is possible for new inodes to be accessed by another thread via the dcache
prior to complete security setup (presently handled by the
post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be
left unlabeled on the disk in the event of a crash. SELinux presently works
around the issue by ensuring that the incore inode security label is
initialized to a special SID that is inaccessible to unprivileged processes
(in accordance with policy), thereby preventing inappropriate access but
potentially causing false denials on legitimate accesses. A simple test
program demonstrates such false denials on SELinux, and the patch solves the
problem. Similar such false denials have been encountered in real
applications.
This patch defines a new inode_init_security LSM hook to obtain the security
attribute to apply to a newly created inode and to set up the incore inode
security state for it, and adds a corresponding hook function implementation
to SELinux.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds endian notations to the SELinux code.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch improves memory use by SELinux by both reducing the avtab node
size and reducing the number of avtab nodes. The memory savings are
substantial, e.g. on a 64-bit system after boot, James Morris reported the
following data for the targeted and strict policies:
#objs objsize kernmem
Targeted:
Before: 237888 40 9.1MB
After: 19968 24 468KB
Strict:
Before: 571680 40 21.81MB
After: 221052 24 5.06MB
The improvement in memory use comes at a cost in the speed of security
server computations of access vectors, but these computations are only
required on AVC cache misses, and performance measurements by James Morris
using a number of benchmarks have shown that the change does not cause any
significant degradation.
Note that a rebuilt policy via an updated policy toolchain
(libsepol/checkpolicy) is required in order to gain the full benefits of
this patch, although some memory savings benefits are immediately applied
even to older policies (in particular, the reduction in avtab node size).
Sources for the updated toolchain are presently available from the
sourceforge CVS tree (http://sourceforge.net/cvs/?group_id=21266), and
tarballs are available from http://www.flux.utah.edu/~sds.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@namei.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
netlink_broadcast users must initialize NETLINK_CB(skb).dst_groups to the
destination group mask for netlink_recvmsg.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
- Remove bogus code for compiling netlink as module
- Add module refcounting support for modules implementing a netlink
protocol
- Add support for autoloading modules that implement a netlink protocol
as soon as someone opens a socket for that protocol
Signed-off-by: Harald Welte <laforge@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch fixes the address length checks in the selinux_socket_connect
hook to be no more restrictive than the underlying ipv4 and ipv6 code;
otherwise, this hook can reject valid connect calls. This patch is in
response to a bug report where an application was calling connect on an
INET6 socket with an address that didn't include the optional scope id and
failing due to these checks.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Implement kernel labeling of the MLS (multilevel security) field of
security contexts for files which have no existing MLS field. This is to
enable upgrades of a system from non-MLS to MLS without performing a full
filesystem relabel including all of the mountpoints, which would be quite
painful for users.
With this patch, with MLS enabled, if a file has no MLS field, the kernel
internally adds an MLS field to the in-core inode (but not to the on-disk
file). This MLS field added is the default for the superblock, allowing
per-mountpoint control over the values via fixed policy or mount options.
This patch has been tested by enabling MLS without relabeling its
filesystem, and seems to be working correctly.
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Stephen Smalley <sds@epoch.ncsc.mil>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Currently selinux_sb_copy_data requires an entire page be allocated to
*orig when the function is called. This "requirement" is based on the fact
that we call copy_page(in_save, nosec_save) and in_save = orig when the
data is not FS_BINARY_MOUNTDATA. This means that if a caller were to call
do_kern_mount with only about 10 bytes of options, they would get passed
here and then we would corrupt PAGE_SIZE - 10 bytes of memory (with all
zeros.)
Currently it appears all in kernel FS's use one page of data so this has
not been a problem. An out of kernel FS did just what is described above
and it would almost always panic shortly after they tried to mount. From
looking else where in the kernel it is obvious that this string of data
must always be null terminated. (See example in do_mount where it always
zeros the last byte.) Thus I suggest we use strcpy in place of copy_page.
In this way we make sure the amount we copy is always less than or equal to
the amount we received and since do_mount is zeroing the last byte this
should be safe for all.
Signed-off-by: Eric Paris <eparis@parisplace.org>
Cc: Stephen Smalley <sds@epoch.ncsc.mil>
Acked-by: James Morris <jmorris@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch,based on sample code by Roland McGrath, adds an execheap
permission check that controls the ability to make the heap executable so
that this can be prevented in almost all cases (the X server is presently
an exception, but this will hopefully be resolved in the future) so that
even programs with execmem permission will need to have the anonymous
memory mapped in order to make it executable.
The only reason that we use a permission check for such restriction (vs.
making it unconditional) is that the X module loader presently needs it; it
could possibly be made unconditional in the future when X is changed.
The policy patch for the execheap permission is available at:
http://pearls.tuxedo-es.org/patches/selinux/policy-execheap.patch
Signed-off-by: Lorenzo Hernandez Garcia-Hierro <lorenzo@gnu.org>
Acked-by: James Morris <jmorris@redhat.com>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds an execstack permission check that controls the ability to
make the main process stack executable so that attempts to make the stack
executable can still be prevented even if the process is allowed the
existing execmem permission in order to e.g. perform runtime code
generation. Note that this does not yet address thread stacks. Note also
that unlike the execmem check, the execstack check is only applied on
mprotect calls, not mmap calls, as the current security_file_mmap hook is
not passed the necessary information presently.
The original author of the code that makes the distinction of the stack
region, is Ingo Molnar, who wrote it within his patch for
/proc/<pid>/maps markers.
(http://marc.theaimsgroup.com/?l=linux-kernel&m=110719881508591&w=2)
The patches also can be found at:
http://pearls.tuxedo-es.org/patches/selinux/policy-execstack.patchhttp://pearls.tuxedo-es.org/patches/selinux/kernel-execstack.patch
policy-execstack.patch is the patch that needs to be applied to the policy in
order to support the execstack permission and exclude it
from general_domain_access within macros/core_macros.te.
kernel-execstack.patch adds such permission to the SELinux code within
the kernel and adds the proper permission check to the selinux_file_mprotect() hook.
Signed-off-by: Lorenzo Hernandez Garcia-Hierro <lorenzo@gnu.org>
Acked-by: James Morris <jmorris@redhat.com>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Add a gfp_mask to audit_log_start() and audit_log(), to reduce the
amount of GFP_ATOMIC allocation -- most of it doesn't need to be
GFP_ATOMIC. Also if the mask includes __GFP_WAIT, then wait up to
60 seconds for the auditd backlog to clear instead of immediately
abandoning the message.
The timeout should probably be made configurable, but for now it'll
suffice that it only happens if auditd is actually running.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
There is a memory leak during mount when SELinux is active and mount
options are specified.
Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com>
Acked-by: Stephen Smalley <sds@epoch.ncsc.mil>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
To retrieve the neighbour tables send RTM_GETNEIGHTBL with the
NLM_F_DUMP flag set. Every neighbour table configuration is
spread over multiple messages to avoid running into message
size limits on systems with many interfaces. The first message
in the sequence transports all not device specific data such as
statistics, configuration, and the default parameter set.
This message is followed by 0..n messages carrying device
specific parameter sets.
Although the ordering should be sufficient, NDTA_NAME can be
used to identify sequences. The initial message can be identified
by checking for NDTA_CONFIG. The device specific messages do
not contain this TLV but have NDTPA_IFINDEX set to the
corresponding interface index.
To change neighbour table attributes, send RTM_SETNEIGHTBL
with NDTA_NAME set. Changeable attribute include NDTA_THRESH[1-3],
NDTA_GC_INTERVAL, and all TLVs in NDTA_PARMS unless marked
otherwise. Device specific parameter sets can be changed by
setting NDTPA_IFINDEX to the interface index of the corresponding
device.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Per Steve Grubb's observation that there are some remaining cases where
avc_audit() directly logs untrusted strings without escaping them, here
is a patch that changes avc_audit() to use audit_log_untrustedstring()
or audit_log_hex() as appropriate. Note that d_name.name is nul-
terminated by d_alloc(), and that sun_path is nul-terminated by
unix_mkname(), so it is not necessary for the AVC to create nul-
terminated copies or to alter audit_log_untrustedstring to take a length
argument. In the case of an abstract name, we use audit_log_hex() with
an explicit length.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
When I added the logging of pid= and comm= back to avc_audit() I
screwed it up. Put it back how it should be.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This patch changes the SELinux AVC to defer logging of paths to the audit
framework upon syscall exit, by saving a reference to the (dentry,vfsmount)
pair in an auxiliary audit item on the current audit context for processing
by audit_log_exit.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
We turned this all off because the 'exe=' was causing deadlocks on
dcache_lock. There's no need to leave the pid and comm out though.
They'll all be logged correctly if full auditing is enabled, but we
should still print them in case auditing _isn't_ enabled.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
It's silly to have to add explicit entries for new userspace messages
as we invent them. Just treat all messages in the user range the same.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This patch should fix the avc_alloc_node() oom condition that Andrew
reported when no policy is loaded in SELinux.
Prior to this patch, when no policy was loaded, the SELinux "security
server" (policy engine) was only returning allowed decisions for the
requested permissions for each access check. This caused the cache to
thrash when trying to use SELinux for real work with no policy loaded
(typically, the no policy loaded state is only for bootstrapping to the
point where we can load an initial policy).
This patch changes the SELinux security server to return the complete
allowed access vector at once, and then to reset the cache after the
initial policy load to flush the initial cache state created during
bootstrapping.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds more messages types to the audit subsystem so that audit
analysis is quicker, intuitive, and more useful.
Signed-off-by: Steve Grubb <sgrubb@redhat.com>
---
I forgot one type in the big patch. I need to add one for user space
originating SE Linux avc messages. This is used by dbus and nscd.
-Steve
---
Updated to 2.6.12-rc4-mm1.
-dwmw2
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Add audit_log_type to allow callers to specify type and pid when logging.
Convert audit_log to wrapper around audit_log_type. Could have
converted all audit_log callers directly, but common case is default
of type AUDIT_KERNEL and pid 0. Update audit_log_start to take type
and pid values when creating a new audit_buffer. Move sequences that
did audit_log_start, audit_log_format, audit_set_type, audit_log_end,
to simply call audit_log_type directly. This obsoletes audit_set_type
and audit_set_pid, so remove them.
Signed-off-by: Chris Wright <chrisw@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
shutdown credential information. It creates a new message type
AUDIT_TERM_INFO, which is used by the audit daemon to query who issued the
shutdown.
It requires the placement of a hook function that gathers the information. The
hook is after the DAC & MAC checks and before the function returns. Racing
threads could overwrite the uid & pid - but they would have to be root and
have policy that allows signalling the audit daemon. That should be a
manageable risk.
The userspace component will be released later in audit 0.7.2. When it
receives the TERM signal, it queries the kernel for shutdown information.
When it receives it, it writes the message and exits. The message looks
like this:
type=DAEMON msg=auditd(1114551182.000) auditd normal halt, sending pid=2650
uid=525, auditd pid=1685
Signed-off-by: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This patch provides finer grained permissions for the audit family of
Netlink sockets under SELinux.
1. We need a way to differentiate between privileged and unprivileged
reads of kernel data maintained by the audit subsystem. The AUDIT_GET
operation is unprivileged: it returns the current status of the audit
subsystem (e.g. whether it's enabled etc.). The AUDIT_LIST operation
however returns a list of the current audit ruleset, which is considered
privileged by the audit folk. To deal with this, a new SELinux
permission has been implemented and applied to the operation:
nlmsg_readpriv, which can be allocated to appropriately privileged
domains. Unprivileged domains would only be allocated nlmsg_read.
2. There is a requirement for certain domains to generate audit events
from userspace. These events need to be collected by the kernel,
collated and transmitted sequentially back to the audit daemon. An
example is user level login, an auditable event under CAPP, where
login-related domains generate AUDIT_USER messages via PAM which are
relayed back to auditd via the kernel. To prevent handing out
nlmsg_write permissions to such domains, a new permission has been
added, nlmsg_relay, which is intended for this type of purpose: data is
passed via the kernel back to userspace but no privileged information is
written to the kernel.
Also, AUDIT_LOGIN messages are now valid only for kernel->user messaging,
so this value has been removed from the SELinux nlmsgtab (which is only
used to check user->kernel messages).
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch removes the sclass argument from ipc_has_perm in the SELinux
module, as it can be obtained from the ipc security structure. The use of
a separate argument was a legacy of the older precondition function
handling in SELinux and is obsolete. Please apply.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The SELinux hooks invoke ipv6_skip_exthdr() with an incorrect
length final argument. However, the length argument turns out
to be superfluous.
I was just reading ipv6_skip_exthdr and it occured to me that we can
get rid of len altogether. The only place where len is used is to
check whether the skb has two bytes for ipv6_opt_hdr. This check
is done by skb_header_pointer/skb_copy_bits anyway.
Now it might appear that we've made the code slower by deferring
the check to skb_copy_bits. However, this check should not trigger
in the common case so this is OK.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
This fixes a deadlock on the dcache lock detected during testing at IBM
by moving the logging of the current executable information from the
SELinux avc_audit function to audit_log_exit (via an audit_log_task_info
helper) for processing upon syscall exit.
For consistency, the patch also removes the logging of other
task-related information from avc_audit, deferring handling to
audit_log_exit instead.
This allows simplification of the avc_audit code, allows the exe
information to be obtained more reliably, always includes the comm
information (useful for scripts), and avoids including bogus task
information for checks performed from irq or softirq.
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch adds SELinux support for the KOBJECT_UEVENT Netlink family, so
that SELinux can apply finer grained controls to it. For example, security
policy for hald can be locked down to the KOBJECT_UEVENT Netlink family
only. Currently, this family simply defaults to the default Netlink socket
class.
Note that some new permission definitions are added to sync with changes in
the core userspace policy package, which auto-generates header files.
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch fixes a bug in the SELinux Netlink message type detection code,
where the wrong constant was being used in a case statement. The incorrect
value is not valid for this class of object so it would not have been
reached, and fallen through to a default handler for all Netlink messages.
Signed-off-by: James Morris <jmorris@redhat.com>
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
Al Viro