13 Dec, 2018
1 commit
-
Historically a lot of these existed because we did not have
a distinction between what was modular code and what was providing
support to modules via EXPORT_SYMBOL and friends. That changed
when we forked out support for the latter into the export.h file.
This means we should be able to reduce the usage of module.h
in code that is obj-y Makefile or bool Kconfig.The advantage in removing such instances is that module.h itself
sources about 15 other headers; adding significantly to what we feed
cpp, and it can obscure what headers we are effectively using.Since module.h might have been the implicit source for init.h
(for __init) and for export.h (for EXPORT_SYMBOL) we consider each
instance for the presence of either and replace as needed.Cc: James Morris
Cc: "Serge E. Hallyn"
Cc: John Johansen
Cc: Mimi Zohar
Cc: Dmitry Kasatkin
Cc: David Howells
Cc: linux-security-module@vger.kernel.org
Cc: linux-integrity@vger.kernel.org
Cc: keyrings@vger.kernel.org
Signed-off-by: Paul Gortmaker
Signed-off-by: James Morris
24 Oct, 2018
1 commit
-
Pull security subsystem updates from James Morris:
"In this patchset, there are a couple of minor updates, as well as some
reworking of the LSM initialization code from Kees Cook (these prepare
the way for ordered stackable LSMs, but are a valuable cleanup on
their own)"* 'next-general' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
LSM: Don't ignore initialization failures
LSM: Provide init debugging infrastructure
LSM: Record LSM name in struct lsm_info
LSM: Convert security_initcall() into DEFINE_LSM()
vmlinux.lds.h: Move LSM_TABLE into INIT_DATA
LSM: Convert from initcall to struct lsm_info
LSM: Remove initcall tracing
LSM: Rename .security_initcall section to .lsm_info
vmlinux.lds.h: Avoid copy/paste of security_init section
LSM: Correctly announce start of LSM initialization
security: fix LSM description location
keys: Fix the use of the C++ keyword "private" in uapi/linux/keyctl.h
seccomp: remove unnecessary unlikely()
security: tomoyo: Fix obsolete function
security/capabilities: remove check for -EINVAL
11 Oct, 2018
6 commits
-
LSM initialization failures have traditionally been ignored. We should
at least WARN when something goes wrong.Signed-off-by: Kees Cook
Reviewed-by: Casey Schaufler
Reviewed-by: John Johansen
Signed-off-by: James Morris -
Booting with "lsm.debug" will report future details on how LSM ordering
decisions are being made.Signed-off-by: Kees Cook
Reviewed-by: Casey Schaufler
Reviewed-by: John Johansen
Reviewed-by: James Morris
Signed-off-by: James Morris -
In preparation for doing more interesting LSM init probing, this converts
the existing initcall system into an explicit call into a function pointer
from a section-collected struct lsm_info array.Signed-off-by: Kees Cook
Reviewed-by: Casey Schaufler
Reviewed-by: James Morris
Reviewed-by: John Johansen
Signed-off-by: James Morris -
This partially reverts commit 58eacfffc417 ("init, tracing: instrument
security and console initcall trace events") since security init calls
are about to no longer resemble regular init calls.Signed-off-by: Kees Cook
Reviewed-by: Casey Schaufler
Reviewed-by: James Morris
Signed-off-by: James Morris -
In preparation for switching from initcall to just a regular set of
pointers in a section, rename the internal section name.Signed-off-by: Kees Cook
Reviewed-by: Casey Schaufler
Reviewed-by: James Morris
Reviewed-by: John Johansen
Signed-off-by: James Morris -
For a while now, the LSM core has said it was "initializED", rather than
"initializING". This adjust the report to be more accurate (i.e. before
this was reported before any LSMs had been initialized.)Signed-off-by: Kees Cook
Reviewed-by: Casey Schaufler
Reviewed-by: James Morris
Reviewed-by: John Johansen
Signed-off-by: James Morris
03 Oct, 2018
1 commit
-
Linus recently observed that if we did not worry about the padding
member in struct siginfo it is only about 48 bytes, and 48 bytes is
much nicer than 128 bytes for allocating on the stack and copying
around in the kernel.The obvious thing of only adding the padding when userspace is
including siginfo.h won't work as there are sigframe definitions in
the kernel that embed struct siginfo.So split siginfo in two; kernel_siginfo and siginfo. Keeping the
traditional name for the userspace definition. While the version that
is used internally to the kernel and ultimately will not be padded to
128 bytes is called kernel_siginfo.The definition of struct kernel_siginfo I have put in include/signal_types.h
A set of buildtime checks has been added to verify the two structures have
the same field offsets.To make it easy to verify the change kernel_siginfo retains the same
size as siginfo. The reduction in size comes in a following change.Signed-off-by: "Eric W. Biederman"
23 Aug, 2018
1 commit
-
Allow the initcall tables to be emitted using relative references that
are only half the size on 64-bit architectures and don't require fixups
at runtime on relocatable kernels.Link: http://lkml.kernel.org/r/20180704083651.24360-5-ard.biesheuvel@linaro.org
Acked-by: James Morris
Acked-by: Sergey Senozhatsky
Acked-by: Petr Mladek
Acked-by: Michael Ellerman
Acked-by: Ingo Molnar
Signed-off-by: Ard Biesheuvel
Cc: Arnd Bergmann
Cc: Benjamin Herrenschmidt
Cc: Bjorn Helgaas
Cc: Catalin Marinas
Cc: James Morris
Cc: Jessica Yu
Cc: Josh Poimboeuf
Cc: Kees Cook
Cc: Nicolas Pitre
Cc: Paul Mackerras
Cc: Russell King
Cc: "Serge E. Hallyn"
Cc: Steven Rostedt
Cc: Thomas Garnier
Cc: Thomas Gleixner
Cc: Will Deacon
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
16 Aug, 2018
2 commits
-
…morris/linux-security
Pull integrity updates from James Morris:
"This adds support for EVM signatures based on larger digests, contains
a new audit record AUDIT_INTEGRITY_POLICY_RULE to differentiate the
IMA policy rules from the IMA-audit messages, addresses two deadlocks
due to either loading or searching for crypto algorithms, and cleans
up the audit messages"* 'next-integrity' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
EVM: fix return value check in evm_write_xattrs()
integrity: prevent deadlock during digsig verification.
evm: Allow non-SHA1 digital signatures
evm: Don't deadlock if a crypto algorithm is unavailable
integrity: silence warning when CONFIG_SECURITYFS is not enabled
ima: Differentiate auditing policy rules from "audit" actions
ima: Do not audit if CONFIG_INTEGRITY_AUDIT is not set
ima: Use audit_log_format() rather than audit_log_string()
ima: Call audit_log_string() rather than logging it untrusted -
Pull security subsystem updates from James Morris:
- kstrdup() return value fix from Eric Biggers
- Add new security_load_data hook to differentiate security checking of
kernel-loaded binaries in the case of there being no associated file
descriptor, from Mimi Zohar.- Add ability to IMA to specify a policy at build-time, rather than
just via command line params or by loading a custom policy, from
Mimi.- Allow IMA and LSMs to prevent sysfs firmware load fallback (e.g. if
using signed firmware), from Mimi.- Allow IMA to deny loading of kexec kernel images, as they cannot be
measured by IMA, from Mimi.* 'next-general' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
security: check for kstrdup() failure in lsm_append()
security: export security_kernel_load_data function
ima: based on policy warn about loading firmware (pre-allocated buffer)
module: replace the existing LSM hook in init_module
ima: add build time policy
ima: based on policy require signed firmware (sysfs fallback)
firmware: add call to LSM hook before firmware sysfs fallback
ima: based on policy require signed kexec kernel images
kexec: add call to LSM hook in original kexec_load syscall
security: define new LSM hook named security_kernel_load_data
MAINTAINERS: remove the outdated "LINUX SECURITY MODULE (LSM) FRAMEWORK" entry
18 Jul, 2018
3 commits
-
This patch aimed to prevent deadlock during digsig verification.The point
of issue - user space utility modprobe and/or it's dependencies (ld-*.so,
libz.so.*, libc-*.so and /lib/modules/ files) that could be used for
kernel modules load during digsig verification and could be signed by
digsig in the same time.First at all, look at crypto_alloc_tfm() work algorithm:
crypto_alloc_tfm() will first attempt to locate an already loaded
algorithm. If that fails and the kernel supports dynamically loadable
modules, it will then attempt to load a module of the same name or alias.
If that fails it will send a query to any loaded crypto manager to
construct an algorithm on the fly.We have situation, when public_key_verify_signature() in case of RSA
algorithm use alg_name to store internal information in order to construct
an algorithm on the fly, but crypto_larval_lookup() will try to use
alg_name in order to load kernel module with same name.1) we can't do anything with crypto module work, since it designed to work
exactly in this way;
2) we can't globally filter module requests for modprobe, since it
designed to work with any requests.In this patch, I propose add an exception for "crypto-pkcs1pad(rsa,*)"
module requests only in case of enabled integrity asymmetric keys support.
Since we don't have any real "crypto-pkcs1pad(rsa,*)" kernel modules for
sure, we are safe to fail such module request from crypto_larval_lookup().
In this way we prevent modprobe execution during digsig verification and
avoid possible deadlock if modprobe and/or it's dependencies also signed
with digsig.Requested "crypto-pkcs1pad(rsa,*)" kernel module name formed by:
1) "pkcs1pad(rsa,%s)" in public_key_verify_signature();
2) "crypto-%s" / "crypto-%s-all" in crypto_larval_lookup().
"crypto-pkcs1pad(rsa," part of request is a constant and unique and could
be used as filter.Signed-off-by: Mikhail Kurinnoi
Signed-off-by: Mimi Zoharinclude/linux/integrity.h | 13 +++++++++++++
security/integrity/digsig_asymmetric.c | 23 +++++++++++++++++++++++
security/security.c | 7 ++++++-
3 files changed, 42 insertions(+), 1 deletion(-) -
lsm_append() should return -ENOMEM if memory allocation failed.
Fixes: d69dece5f5b6 ("LSM: Add /sys/kernel/security/lsm")
Signed-off-by: Eric Biggers
Signed-off-by: James Morris -
The firmware_loader can be built as a loadable module, which now
fails when CONFIG_SECURITY is enabled, because a call to the
security_kernel_load_data() function got added, and this is
not exported to modules:ERROR: "security_kernel_load_data" [drivers/base/firmware_loader/firmware_class.ko] undefined!
Add an EXPORT_SYMBOL_GPL() to make it available here.
Fixes: 6e852651f28e ("firmware: add call to LSM hook before firmware sysfs fallback")
Signed-off-by: Arnd Bergmann
Signed-off-by: James Morris
17 Jul, 2018
2 commits
-
The original kexec_load syscall can not verify file signatures, nor can
the kexec image be measured. Based on policy, deny the kexec_load
syscall.Signed-off-by: Mimi Zohar
Cc: Eric Biederman
Cc: Kees Cook
Reviewed-by: Kees Cook
Signed-off-by: James Morris -
Differentiate between the kernel reading a file specified by userspace
from the kernel loading a buffer containing data provided by userspace.
This patch defines a new LSM hook named security_kernel_load_data().Signed-off-by: Mimi Zohar
Cc: Eric Biederman
Cc: Luis R. Rodriguez
Cc: Kees Cook
Cc: Casey Schaufler
Acked-by: Serge Hallyn
Acked-by: Kees Cook
Signed-off-by: James Morris
12 Jul, 2018
2 commits
-
Acked-by: Linus Torvalds
Signed-off-by: Al Viro -
Acked-by: Linus Torvalds
Signed-off-by: Al Viro
05 May, 2018
1 commit
-
Right now the LSM labels for socketpairs are always uninitialized,
since there is no security hook for the socketpair() syscall. This
patch adds the required hooks so LSMs can properly label socketpairs.
This allows SO_PEERSEC to return useful information on those sockets.Note that the behavior of socketpair() can be emulated by creating a
listener socket, connecting to it, and then discarding the initial
listener socket. With this workaround, SO_PEERSEC would return the
caller's security context. However, with socketpair(), the uninitialized
context is returned unconditionally. This is unexpected and makes
socketpair() less useful in situations where the security context is
crucial to the application.With the new socketpair-hook this disparity can be solved by making
socketpair() return the expected security context.Acked-by: Serge Hallyn
Signed-off-by: Tom Gundersen
Signed-off-by: David Herrmann
Signed-off-by: James Morris
11 Apr, 2018
1 commit
-
Pull tracing updates from Steven Rostedt:
"New features:- Tom Zanussi's extended histogram work.
This adds the synthetic events to have histograms from multiple
event data Adds triggers "onmatch" and "onmax" to call the
synthetic events Several updates to the histogram code from this- Allow way to nest ring buffer calls in the same context
- Allow absolute time stamps in ring buffer
- Rewrite of filter code parsing based on Al Viro's suggestions
- Setting of trace_clock to global if TSC is unstable (on boot)
- Better OOM handling when allocating large ring buffers
- Added initcall tracepoints (consolidated initcall_debug code with
them)And other various fixes and clean ups"
* tag 'trace-v4.17' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (68 commits)
init: Have initcall_debug still work without CONFIG_TRACEPOINTS
init, tracing: Have printk come through the trace events for initcall_debug
init, tracing: instrument security and console initcall trace events
init, tracing: Add initcall trace events
tracing: Add rcu dereference annotation for test func that touches filter->prog
tracing: Add rcu dereference annotation for filter->prog
tracing: Fixup logic inversion on setting trace_global_clock defaults
tracing: Hide global trace clock from lockdep
ring-buffer: Add set/clear_current_oom_origin() during allocations
ring-buffer: Check if memory is available before allocation
lockdep: Add print_irqtrace_events() to __warn
vsprintf: Do not preprocess non-dereferenced pointers for bprintf (%px and %pK)
tracing: Uninitialized variable in create_tracing_map_fields()
tracing: Make sure variable string fields are NULL-terminated
tracing: Add action comparisons when testing matching hist triggers
tracing: Don't add flag strings when displaying variable references
tracing: Fix display of hist trigger expressions containing timestamps
ftrace: Drop a VLA in module_exists()
tracing: Mention trace_clock=global when warning about unstable clocks
tracing: Default to using trace_global_clock if sched_clock is unstable
...
08 Apr, 2018
2 commits
-
…morris/linux-security
Pull integrity updates from James Morris:
"A mixture of bug fixes, code cleanup, and continues to close
IMA-measurement, IMA-appraisal, and IMA-audit gaps.Also note the addition of a new cred_getsecid LSM hook by Matthew
Garrett:For IMA purposes, we want to be able to obtain the prepared secid
in the bprm structure before the credentials are committed. Add a
cred_getsecid hook that makes this possible.which is used by a new CREDS_CHECK target in IMA:
In ima_bprm_check(), check with both the existing process
credentials and the credentials that will be committed when the new
process is started. This will not change behaviour unless the
system policy is extended to include CREDS_CHECK targets -
BPRM_CHECK will continue to check the same credentials that it did
previously"* 'next-integrity' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
ima: Fallback to the builtin hash algorithm
ima: Add smackfs to the default appraise/measure list
evm: check for remount ro in progress before writing
ima: Improvements in ima_appraise_measurement()
ima: Simplify ima_eventsig_init()
integrity: Remove unused macro IMA_ACTION_RULE_FLAGS
ima: drop vla in ima_audit_measurement()
ima: Fix Kconfig to select TPM 2.0 CRB interface
evm: Constify *integrity_status_msg[]
evm: Move evm_hmac and evm_hash from evm_main.c to evm_crypto.c
fuse: define the filesystem as untrusted
ima: fail signature verification based on policy
ima: clear IMA_HASH
ima: re-evaluate files on privileged mounted filesystems
ima: fail file signature verification on non-init mounted filesystems
IMA: Support using new creds in appraisal policy
security: Add a cred_getsecid hook -
Pull general security layer updates from James Morris:
- Convert security hooks from list to hlist, a nice cleanup, saving
about 50% of space, from Sargun Dhillon.- Only pass the cred, not the secid, to kill_pid_info_as_cred and
security_task_kill (as the secid can be determined from the cred),
from Stephen Smalley.- Close a potential race in kernel_read_file(), by making the file
unwritable before calling the LSM check (vs after), from Kees Cook.* 'next-general' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
security: convert security hooks to use hlist
exec: Set file unwritable before LSM check
usb, signal, security: only pass the cred, not the secid, to kill_pid_info_as_cred and security_task_kill
07 Apr, 2018
1 commit
-
Pull SELinux updates from Paul Moore:
"A bigger than usual pull request for SELinux, 13 patches (lucky!)
along with a scary looking diffstat.Although if you look a bit closer, excluding the usual minor
tweaks/fixes, there are really only two significant changes in this
pull request: the addition of proper SELinux access controls for SCTP
and the encapsulation of a lot of internal SELinux state.The SCTP changes are the result of a multi-month effort (maybe even a
year or longer?) between the SELinux folks and the SCTP folks to add
proper SELinux controls. A special thanks go to Richard for seeing
this through and keeping the effort moving forward.The state encapsulation work is a bit of janitorial work that came out
of some early work on SELinux namespacing. The question of namespacing
is still an open one, but I believe there is some real value in the
encapsulation work so we've split that out and are now sending that up
to you"* tag 'selinux-pr-20180403' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/selinux:
selinux: wrap AVC state
selinux: wrap selinuxfs state
selinux: fix handling of uninitialized selinux state in get_bools/classes
selinux: Update SELinux SCTP documentation
selinux: Fix ltp test connect-syscall failure
selinux: rename the {is,set}_enforcing() functions
selinux: wrap global selinux state
selinux: fix typo in selinux_netlbl_sctp_sk_clone declaration
selinux: Add SCTP support
sctp: Add LSM hooks
sctp: Add ip option support
security: Add support for SCTP security hooks
netlabel: If PF_INET6, check sk_buff ip header version
06 Apr, 2018
1 commit
-
Trace events have been added around the initcall functions defined in
init/main.c. But console and security have their own initcalls. This adds
the trace events associated for those initcall functions.Link: http://lkml.kernel.org/r/1521765208.19745.2.camel@polymtl.ca
Cc: Ingo Molnar
Cc: Peter Zijlstra
Signed-off-by: Abderrahmane Benbachir
Signed-off-by: Steven Rostedt (VMware)
31 Mar, 2018
1 commit
-
This changes security_hook_heads to use hlist_heads instead of
the circular doubly-linked list heads. This should cut down
the size of the struct by about half.In addition, it allows mutation of the hooks at the tail of the
callback list without having to modify the head. The longer-term
purpose of this is to enable making the heads read only.Signed-off-by: Sargun Dhillon
Reviewed-by: Tetsuo Handa
Acked-by: Casey Schaufler
Signed-off-by: James Morris
23 Mar, 2018
4 commits
-
For IMA purposes, we want to be able to obtain the prepared secid in the
bprm structure before the credentials are committed. Add a cred_getsecid
hook that makes this possible.Signed-off-by: Matthew Garrett
Acked-by: Paul Moore
Cc: Paul Moore
Cc: Stephen Smalley
Cc: Casey Schaufler
Signed-off-by: Mimi Zohar -
All of the implementations of security hooks that take msg_queue only
access q_perm the struct kern_ipc_perm member. This means the
dependencies of the msg_queue security hooks can be simplified by
passing the kern_ipc_perm member of msg_queue.Making this change will allow struct msg_queue to become private to
ipc/msg.c.Signed-off-by: "Eric W. Biederman"
-
All of the implementations of security hooks that take shmid_kernel only
access shm_perm the struct kern_ipc_perm member. This means the
dependencies of the shm security hooks can be simplified by passing
the kern_ipc_perm member of shmid_kernel..Making this change will allow struct shmid_kernel to become private to ipc/shm.c.
Signed-off-by: "Eric W. Biederman"
-
All of the implementations of security hooks that take sem_array only
access sem_perm the struct kern_ipc_perm member. This means the
dependencies of the sem security hooks can be simplified by passing
the kern_ipc_perm member of sem_array.Making this change will allow struct sem and struct sem_array
to become private to ipc/sem.c.Signed-off-by: "Eric W. Biederman"
07 Mar, 2018
1 commit
-
…ed and security_task_kill
commit d178bc3a708f39cbfefc3fab37032d3f2511b4ec ("user namespace: usb:
make usb urbs user namespace aware (v2)") changed kill_pid_info_as_uid
to kill_pid_info_as_cred, saving and passing a cred structure instead of
uids. Since the secid can be obtained from the cred, drop the secid fields
from the usb_dev_state and async structures, and drop the secid argument to
kill_pid_info_as_cred. Replace the secid argument to security_task_kill
with the cred. Update SELinux, Smack, and AppArmor to use the cred, which
avoids the need for Smack and AppArmor to use a secid at all in this hook.
Further changes to Smack might still be required to take full advantage of
this change, since it should now be possible to perform capability
checking based on the supplied cred. The changes to Smack and AppArmor
have only been compile-tested.Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
Acked-by: Paul Moore <paul@paul-moore.com>
Acked-by: Casey Schaufler <casey@schaufler-ca.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: James Morris <james.morris@microsoft.com>
23 Feb, 2018
1 commit
-
The SCTP security hooks are explained in:
Documentation/security/LSM-sctp.rstSigned-off-by: Richard Haines
Signed-off-by: Paul Moore
20 Oct, 2017
1 commit
-
Introduce several LSM hooks for the syscalls that will allow the
userspace to access to eBPF object such as eBPF programs and eBPF maps.
The security check is aimed to enforce a per object security protection
for eBPF object so only processes with the right priviliges can
read/write to a specific map or use a specific eBPF program. Besides
that, a general security hook is added before the multiplexer of bpf
syscall to check the cmd and the attribute used for the command. The
actual security module can decide which command need to be checked and
how the cmd should be checked.Signed-off-by: Chenbo Feng
Acked-by: James Morris
Signed-off-by: David S. Miller
13 Sep, 2017
1 commit
-
Pull selinux updates from Paul Moore:
"A relatively quiet period for SELinux, 11 patches with only two/three
having any substantive changes.These noteworthy changes include another tweak to the NNP/nosuid
handling, per-file labeling for cgroups, and an object class fix for
AF_UNIX/SOCK_RAW sockets; the rest of the changes are minor tweaks or
administrative updates (Stephen's email update explains the file
explosion in the diffstat).Everything passes the selinux-testsuite"
[ Also a couple of small patches from the security tree from Tetsuo
Handa for Tomoyo and LSM cleanup. The separation of security policy
updates wasn't all that clean - Linus ]* tag 'selinux-pr-20170831' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/selinux:
selinux: constify nf_hook_ops
selinux: allow per-file labeling for cgroupfs
lsm_audit: update my email address
selinux: update my email address
MAINTAINERS: update the NetLabel and Labeled Networking information
selinux: use GFP_NOWAIT in the AVC kmem_caches
selinux: Generalize support for NNP/nosuid SELinux domain transitions
selinux: genheaders should fail if too many permissions are defined
selinux: update the selinux info in MAINTAINERS
credits: update Paul Moore's info
selinux: Assign proper class to PF_UNIX/SOCK_RAW sockets
tomoyo: Update URLs in Documentation/admin-guide/LSM/tomoyo.rst
LSM: Remove security_task_create() hook.
02 Aug, 2017
1 commit
-
This removes the bprm_secureexec hook since the logic has been folded into
the bprm_set_creds hook for all LSMs now.Cc: Eric W. Biederman
Signed-off-by: Kees Cook
Reviewed-by: John Johansen
Acked-by: James Morris
Acked-by: Serge Hallyn
18 Jul, 2017
1 commit
-
Since commit a79be238600d1a03 ("selinux: Use task_alloc hook rather than
task_create hook") changed to use task_alloc hook, task_create hook is
no longer used.Signed-off-by: Tetsuo Handa
Signed-off-by: James Morris
10 Jun, 2017
1 commit
-
When an NFSv4 client performs a mount operation, it first mounts the
NFSv4 root and then does path walk to the exported path and performs a
submount on that, cloning the security mount options from the root's
superblock to the submount's superblock in the process.Unless the NFS server has an explicit fsid=0 export with the
"security_label" option, the NFSv4 root superblock will not have
SBLABEL_MNT set, and neither will the submount superblock after cloning
the security mount options. As a result, setxattr's of security labels
over NFSv4.2 will fail. In a similar fashion, NFSv4.2 mounts mounted
with the context= mount option will not show the correct labels because
the nfs_server->caps flags of the cloned superblock will still have
NFS_CAP_SECURITY_LABEL set.Allowing the NFSv4 client to enable or disable SECURITY_LSM_NATIVE_LABELS
behavior will ensure that the SBLABEL_MNT flag has the correct value
when the client traverses from an exported path without the
"security_label" option to one with the "security_label" option and
vice versa. Similarly, checking to see if SECURITY_LSM_NATIVE_LABELS is
set upon return from security_sb_clone_mnt_opts() and clearing
NFS_CAP_SECURITY_LABEL if necessary will allow the correct labels to
be displayed for NFSv4.2 mounts mounted with the context= mount option.Resolves: https://github.com/SELinuxProject/selinux-kernel/issues/35
Signed-off-by: Scott Mayhew
Reviewed-by: Stephen Smalley
Tested-by: Stephen Smalley
Signed-off-by: Paul Moore
24 May, 2017
3 commits
-
Allocate and free a security context when creating and destroying a MAD
agent. This context is used for controlling access to PKeys and sending
and receiving SMPs.When sending or receiving a MAD check that the agent has permission to
access the PKey for the Subnet Prefix of the port.During MAD and snoop agent registration for SMI QPs check that the
calling process has permission to access the manage the subnet and
register a callback with the LSM to be notified of policy changes. When
notificaiton of a policy change occurs recheck permission and set a flag
indicating sending and receiving SMPs is allowed.When sending and receiving MADs check that the agent has access to the
SMI if it's on an SMI QP. Because security policy can change it's
possible permission was allowed when creating the agent, but no longer
is.Signed-off-by: Daniel Jurgens
Acked-by: Doug Ledford
[PM: remove the LSM hook init code]
Signed-off-by: Paul Moore -
Add a generic notificaiton mechanism in the LSM. Interested consumers
can register a callback with the LSM and security modules can produce
events.Because access to Infiniband QPs are enforced in the setup phase of a
connection security should be enforced again if the policy changes.
Register infiniband devices for policy change notification and check all
QPs on that device when the notification is received.Add a call to the notification mechanism from SELinux when the AVC
cache changes or setenforce is cleared.Signed-off-by: Daniel Jurgens
Acked-by: James Morris
Acked-by: Doug Ledford
Signed-off-by: Paul Moore -
Add new LSM hooks to allocate and free security contexts and check for
permission to access a PKey.Allocate and free a security context when creating and destroying a QP.
This context is used for controlling access to PKeys.When a request is made to modify a QP that changes the port, PKey index,
or alternate path, check that the QP has permission for the PKey in the
PKey table index on the subnet prefix of the port. If the QP is shared
make sure all handles to the QP also have access.Store which port and PKey index a QP is using. After the reset to init
transition the user can modify the port, PKey index and alternate path
independently. So port and PKey settings changes can be a merge of the
previous settings and the new ones.In order to maintain access control if there are PKey table or subnet
prefix change keep a list of all QPs are using each PKey index on
each port. If a change occurs all QPs using that device and port must
have access enforced for the new cache settings.These changes add a transaction to the QP modify process. Association
with the old port and PKey index must be maintained if the modify fails,
and must be removed if it succeeds. Association with the new port and
PKey index must be established prior to the modify and removed if the
modify fails.1. When a QP is modified to a particular Port, PKey index or alternate
path insert that QP into the appropriate lists.2. Check permission to access the new settings.
3. If step 2 grants access attempt to modify the QP.
4a. If steps 2 and 3 succeed remove any prior associations.
4b. If ether fails remove the new setting associations.
If a PKey table or subnet prefix changes walk the list of QPs and
check that they have permission. If not send the QP to the error state
and raise a fatal error event. If it's a shared QP make sure all the
QPs that share the real_qp have permission as well. If the QP that
owns a security structure is denied access the security structure is
marked as such and the QP is added to an error_list. Once the moving
the QP to error is complete the security structure mark is cleared.Maintaining the lists correctly turns QP destroy into a transaction.
The hardware driver for the device frees the ib_qp structure, so while
the destroy is in progress the ib_qp pointer in the ib_qp_security
struct is undefined. When the destroy process begins the ib_qp_security
structure is marked as destroying. This prevents any action from being
taken on the QP pointer. After the QP is destroyed successfully it
could still listed on an error_list wait for it to be processed by that
flow before cleaning up the structure.If the destroy fails the QPs port and PKey settings are reinserted into
the appropriate lists, the destroying flag is cleared, and access control
is enforced, in case there were any cache changes during the destroy
flow.To keep the security changes isolated a new file is used to hold security
related functionality.Signed-off-by: Daniel Jurgens
Acked-by: Doug Ledford
[PM: merge fixup in ib_verbs.h and uverbs_cmd.c]
Signed-off-by: Paul Moore
