Pull namespace updates from Eric Biederman:
"There is a lot here. A lot of these changes result in subtle user
visible differences in kernel behavior. I don't expect anything will
care but I will revert/fix things immediately if any regressions show
up.
From Seth Forshee there is a continuation of the work to make the vfs
ready for unpriviled mounts. We had thought the previous changes
prevented the creation of files outside of s_user_ns of a filesystem,
but it turns we missed the O_CREAT path. Ooops.
Pavel Tikhomirov and Oleg Nesterov worked together to fix a long
standing bug in the implemenation of PR_SET_CHILD_SUBREAPER where only
children that are forked after the prctl are considered and not
children forked before the prctl. The only known user of this prctl
systemd forks all children after the prctl. So no userspace
regressions will occur. Holding earlier forked children to the same
rules as later forked children creates a semantic that is sane enough
to allow checkpoing of processes that use this feature.
There is a long delayed change by Nikolay Borisov to limit inotify
instances inside a user namespace.
Michael Kerrisk extends the API for files used to maniuplate
namespaces with two new trivial ioctls to allow discovery of the
hierachy and properties of namespaces.
Konstantin Khlebnikov with the help of Al Viro adds code that when a
network namespace exits purges it's sysctl entries from the dcache. As
in some circumstances this could use a lot of memory.
Vivek Goyal fixed a bug with stacked filesystems where the permissions
on the wrong inode were being checked.
I continue previous work on ptracing across exec. Allowing a file to
be setuid across exec while being ptraced if the tracer has enough
credentials in the user namespace, and if the process has CAP_SETUID
in it's own namespace. Proc files for setuid or otherwise undumpable
executables are now owned by the root in the user namespace of their
mm. Allowing debugging of setuid applications in containers to work
better.
A bug I introduced with permission checking and automount is now
fixed. The big change is to mark the mounts that the kernel initiates
as a result of an automount. This allows the permission checks in sget
to be safely suppressed for this kind of mount. As the permission
check happened when the original filesystem was mounted.
Finally a special case in the mount namespace is removed preventing
unbounded chains in the mount hash table, and making the semantics
simpler which benefits CRIU.
The vfs fix along with related work in ima and evm I believe makes us
ready to finish developing and merge fully unprivileged mounts of the
fuse filesystem. The cleanups of the mount namespace makes discussing
how to fix the worst case complexity of umount. The stacked filesystem
fixes pave the way for adding multiple mappings for the filesystem
uids so that efficient and safer containers can be implemented"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace:
proc/sysctl: Don't grab i_lock under sysctl_lock.
vfs: Use upper filesystem inode in bprm_fill_uid()
proc/sysctl: prune stale dentries during unregistering
mnt: Tuck mounts under others instead of creating shadow/side mounts.
prctl: propagate has_child_subreaper flag to every descendant
introduce the walk_process_tree() helper
nsfs: Add an ioctl() to return owner UID of a userns
fs: Better permission checking for submounts
exit: fix the setns() && PR_SET_CHILD_SUBREAPER interaction
vfs: open() with O_CREAT should not create inodes with unknown ids
nsfs: Add an ioctl() to return the namespace type
proc: Better ownership of files for non-dumpable tasks in user namespaces
exec: Remove LSM_UNSAFE_PTRACE_CAP
exec: Test the ptracer's saved cred to see if the tracee can gain caps
exec: Don't reset euid and egid when the tracee has CAP_SETUID
inotify: Convert to using per-namespace limits
With previous changes every location that tests for
LSM_UNSAFE_PTRACE_CAP also tests for LSM_UNSAFE_PTRACE making the
LSM_UNSAFE_PTRACE_CAP redundant, so remove it.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Moving the use of fqname to later allows learning profiles to be based
on the fqname request instead of just the hname. It also allows cleaning
up some of the name parsing and lookup by allowing the use of
the fqlookupn_profile() lib fn.
Signed-off-by: John Johansen <john.johansen@canonical.com>
The aad macro can replace aad strings when it is not intended to. Switch
to a fn macro so it is only applied when intended.
Also at the same time cleanup audit_data initialization by putting
common boiler plate behind a macro, and dropping the gfp_t parameter
which will become useless.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Having ops be an integer that is an index into an op name table is
awkward and brittle. Every op change requires an edit for both the
op constant and a string in the table. Instead switch to using const
strings directly, eliminating the need for the table that needs to
be kept in sync.
Signed-off-by: John Johansen <john.johansen@canonical.com>
When possible its better to name a learning profile after the missing
profile in question. This allows for both more informative names and
for profile reuse.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Policy namespaces will be diverging from profile management and
expanding so put it in its own file.
Signed-off-by: John Johansen <john.johansen@canonical.com>
After a policy replacement, the task cred may be out of date and need
to be updated. However change_hat is using the stale profiles from
the out of date cred resulting in either: a stale profile being applied
or, incorrect failure when searching for a hat profile as it has been
migrated to the new parent profile.
Fixes: 01e2b670aa (failure to find hat)
Fixes: 898127c34e (stale policy being applied)
Bugzilla: https://bugzilla.suse.com/show_bug.cgi?id=1000287
Cc: stable@vger.kernel.org
Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: James Morris <james.l.morris@oracle.com>
The target profile name was not being correctly audited in a few
cases because the target variable was not being set and gotos
passed the code to set it at apply:
Since it is always based on new_profile just drop the target var
and conditionally report based on new_profile.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
The current behavior is confusing as it causes exec failures to report
the executable is missing instead of identifying that apparmor
caused the failure.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
Instead of using a vector of security operations
with explicit, special case stacking of the capability
and yama hooks use lists of hooks with capability and
yama hooks included as appropriate.
The security_operations structure is no longer required.
Instead, there is a union of the function pointers that
allows all the hooks lists to use a common mechanism for
list management while retaining typing. Each module
supplies an array describing the hooks it provides instead
of a sparsely populated security_operations structure.
The description includes the element that gets put on
the hook list, avoiding the issues surrounding individual
element allocation.
The method for registering security modules is changed to
reflect the information available. The method for removing
a module, currently only used by SELinux, has also changed.
It should be generic now, however if there are potential
race conditions based on ordering of hook removal that needs
to be addressed by the calling module.
The security hooks are called from the lists and the first
failure is returned.
Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <keescook@chromium.org>
Acked-by: Paul Moore <paul@paul-moore.com>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
Acked-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: James Morris <james.l.morris@oracle.com>
Since seccomp transitions between threads requires updates to the
no_new_privs flag to be atomic, the flag must be part of an atomic flag
set. This moves the nnp flag into a separate task field, and introduces
accessors.
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Reviewed-by: Andy Lutomirski <luto@amacapital.net>
Unless task == current ptrace_parent(task) is not safe even under
rcu_read_lock() and most of the current users are not right.
So may_change_ptraced_domain(task) looks wrong as well. However it
is always called with task == current so the code is actually fine.
Remove this argument to make this fact clear.
Note: perhaps we should simply kill ptrace_parent(), it buys almost
nothing. And it is obviously racy, perhaps this should be fixed.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>
Mediation is based off of the cred but auditing includes the current
task which may not be related to the actual request.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Allow emulating the default profile behavior from boot, by allowing
loading of a profile in the unconfined state into a new NS.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
ns->unconfined is being used read side without locking, nor rcu but is
being updated when a namespace is removed. This works for the root ns
which is never removed but has a race window and can cause failures when
children namespaces are removed.
Also ns and ns->unconfined have a circular refcounting dependency that
is problematic and must be broken. Currently this is done incorrectly
when the namespace is destroyed.
Fix this by forward referencing unconfined via the replacedby infrastructure
instead of directly updating the ns->unconfined pointer.
Remove the circular refcount dependency by making the ns and its unconfined
profile share the same refcount.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Seth Arnold <seth.arnold@canonical.com>
remove the use of replaced by chaining and move to profile invalidation
and lookup to handle task replacement.
Replacement chaining can result in large chains of profiles being pinned
in memory when one profile in the chain is use. With implicit labeling
this will be even more of a problem, so move to a direct lookup method.
Signed-off-by: John Johansen <john.johansen@canonical.com>
When policy specifies a transition to a profile that is not currently
loaded, it result in exec being denied. However the failure is not being
audited correctly because the audit code is treating this as an allowed
permission and thus not reporting it.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-By: Steve Beattie <sbeattie@ubuntu.com>
Cc: John Johansen <john.johansen@canonical.com>
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Add support for AppArmor to explicitly fail requested domain transitions
if NO_NEW_PRIVS is set and the task is not unconfined.
Transitions from unconfined are still allowed because this always results
in a reduction of privileges.
Acked-by: Eric Paris <eparis@redhat.com>
Signed-off-by: Will Drewry <wad@chromium.org>
Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: Andy Lutomirski <luto@amacapital.net>
v18: new acked-by, new description
Signed-off-by: James Morris <james.l.morris@oracle.com>
With this change, calling
prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)
disables privilege granting operations at execve-time. For example, a
process will not be able to execute a setuid binary to change their uid
or gid if this bit is set. The same is true for file capabilities.
Additionally, LSM_UNSAFE_NO_NEW_PRIVS is defined to ensure that
LSMs respect the requested behavior.
To determine if the NO_NEW_PRIVS bit is set, a task may call
prctl(PR_GET_NO_NEW_PRIVS, 0, 0, 0, 0);
It returns 1 if set and 0 if it is not set. If any of the arguments are
non-zero, it will return -1 and set errno to -EINVAL.
(PR_SET_NO_NEW_PRIVS behaves similarly.)
This functionality is desired for the proposed seccomp filter patch
series. By using PR_SET_NO_NEW_PRIVS, it allows a task to modify the
system call behavior for itself and its child tasks without being
able to impact the behavior of a more privileged task.
Another potential use is making certain privileged operations
unprivileged. For example, chroot may be considered "safe" if it cannot
affect privileged tasks.
Note, this patch causes execve to fail when PR_SET_NO_NEW_PRIVS is
set and AppArmor is in use. It is fixed in a subsequent patch.
Signed-off-by: Andy Lutomirski <luto@amacapital.net>
Signed-off-by: Will Drewry <wad@chromium.org>
Acked-by: Eric Paris <eparis@redhat.com>
Acked-by: Kees Cook <keescook@chromium.org>
v18: updated change desc
v17: using new define values as per 3.4
Signed-off-by: James Morris <james.l.morris@oracle.com>
Fix failure in aa_change_onexec api when the request is made from a confined
task. This failure was caused by two problems
The AA_MAY_ONEXEC perm was not being mapped correctly for this case.
The executable name was being checked as second time instead of using the
requested onexec profile name, which may not be the same as the exec
profile name. This mistake can not be exploited to grant extra permission
because of the above flaw where the ONEXEC permission was not being mapped
so it will not be granted.
BugLink: http://bugs.launchpad.net/bugs/963756
Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: James Morris <james.l.morris@oracle.com>
Move the path name lookup failure messages into the main path name lookup
routine, as the information is useful in more than just aa_path_perm.
Also rename aa_get_name to aa_path_name as it is not getting a reference
counted object with a corresponding put fn.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Acked-by: Kees Cook <kees@ubuntu.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6: (54 commits)
tpm_nsc: Fix bug when loading multiple TPM drivers
tpm: Move tpm_tis_reenable_interrupts out of CONFIG_PNP block
tpm: Fix compilation warning when CONFIG_PNP is not defined
TOMOYO: Update kernel-doc.
tpm: Fix a typo
tpm_tis: Probing function for Intel iTPM bug
tpm_tis: Fix the probing for interrupts
tpm_tis: Delay ACPI S3 suspend while the TPM is busy
tpm_tis: Re-enable interrupts upon (S3) resume
tpm: Fix display of data in pubek sysfs entry
tpm_tis: Add timeouts sysfs entry
tpm: Adjust interface timeouts if they are too small
tpm: Use interface timeouts returned from the TPM
tpm_tis: Introduce durations sysfs entry
tpm: Adjust the durations if they are too small
tpm: Use durations returned from TPM
TOMOYO: Enable conditional ACL.
TOMOYO: Allow using argv[]/envp[] of execve() as conditions.
TOMOYO: Allow using executable's realpath and symlink's target as conditions.
TOMOYO: Allow using owner/group etc. of file objects as conditions.
...
Fix up trivial conflict in security/tomoyo/realpath.c
The pointer returned from tracehook_tracer_task() is only valid inside
the rcu_read_lock. However the tracer pointer obtained is being passed
to aa_may_ptrace outside of the rcu_read_lock critical section.
Mover the aa_may_ptrace test into the rcu_read_lock critical section, to
fix this.
Kernels affected: 2.6.36 - 3.0
Reported-by: Oleg Nesterov <oleg@redhat.com>
Cc: stable@kernel.org
Signed-off-by: John Johansen <john.johansen@canonical.com>
tracehook.h is on the way out. Rename tracehook_tracer_task() to
ptrace_parent() and move it from tracehook.h to ptrace.h.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: John Johansen <john.johansen@canonical.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
AppArmor routines for controling domain transitions, which can occur at
exec or through self directed change_profile/change_hat calls.
Unconfined tasks are checked at exec against the profiles in the confining
profile namespace to determine if a profile should be attached to the task.
Confined tasks execs are controlled by the profile which provides rules
determining which execs are allowed and if so which profiles should be
transitioned to.
Self directed domain transitions allow a task to request transition
to a given profile. If the transition is allowed then the profile will
be applied, either immeditately or at exec time depending on the request.
Immeditate self directed transitions have several security limitations
but have uses in setting up stub transition profiles and other limited
cases.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: James Morris <jmorris@namei.org>
Linus Torvalds
Eric W. Biederman
John Johansen
Casey Schaufler
Kees Cook
Oleg Nesterov
Al Viro
Andy Lutomirski
Tejun Heo
James Morris