linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing
@ 2019-07-21 21:31 Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 01/10] fs,security: Add a new file access type: MAY_CHROOT Mickaël Salaün
                   ` (9 more replies)
  0 siblings, 10 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

Hi,

This tenth series mainly replace the previous [1] inode map
implementation with a hash map, which assure uniqueness of keys, improve
performance, and switch to arbitrary value size.  The inode and map
lifetime are now handled by LSM hooks.  The previous subtype is replaced
with the already existing expected attach type and a new expected attach
triggers field.

Landlock is a stackable LSM [4] intended to be used as a low-level
framework to build custom access-control systems or safe endpoint
security agents.  There is two types of Landlock hooks: FS_WALK and
FS_PICK.  Each of them accepts a dedicated eBPF program, called a
Landlock program.  The set of actions on a file is well defined (e.g.
read, write, ioctl, append, lock, mount...) taking inspiration from the
major Linux LSMs and some other access-controls like Capsicum.

The example patch show how a file system access control can be built
based on a list of denied files and directories.  From a security point
of view, it may be preferable to use a whitelist instead of a blacklist,
but this series only enable to match a specific list of files.  Bringing
back a way to evaluate a path is planned for a future dedicated series,
once this base Landlock framework is merged.  I may take inspiration
from the LOOKUP_BENEATH approach [5], but from an eBPF point of view.

The documentation patch contains some kernel documentation and
explanations on how to use Landlock.  The compiled documentation and
some talks can be found here: https://landlock.io
This patch series can be found in a Git repository here:
https://github.com/landlock-lsm/linux/commits/landlock-v10

This is the first step of the roadmap discussed at LPC [2].  While the
intended final goal is to allow unprivileged users to use Landlock, this
series allows only a process with global CAP_SYS_ADMIN to load and
enforce a rule.  This may help to get feedback and avoid unexpected
behaviors.

This series can be applied on top of bpf-next, commit 88091ff56b71
("selftests, bpf: Add test for veth native XDP").  This can be tested
with CONFIG_SECCOMP_FILTER and CONFIG_SECURITY_LANDLOCK.  I would really
appreciate constructive comments on the design and the code.


# Landlock LSM

The goal of this new Linux Security Module (LSM) called Landlock is to
allow any process, including unprivileged ones, to create powerful
security sandboxes comparable to XNU Sandbox or OpenBSD Pledge (which
could be implemented with Landlock).  This kind of sandbox is expected
to help mitigate the security impact of bugs or unexpected/malicious
behaviors in user-space applications.

The approach taken is to add the minimum amount of code while still
allowing the user-space application to create quite complex access
rules.  A dedicated security policy language such as the one used by
SELinux, AppArmor and other major LSMs involves a lot of code and is
usually permitted to only a trusted user (i.e. root).  On the contrary,
eBPF programs already exist and are designed to be safely loaded by
unprivileged user-space.

This design does not seem too intrusive but is flexible enough to allow
a powerful sandbox mechanism accessible by any process on Linux. The use
of seccomp and Landlock is more suitable with the help of a user-space
library (e.g.  libseccomp) that could help to specify a high-level
language to express a security policy instead of raw eBPF programs.
Moreover, thanks to the LLVM front-end, it is quite easy to write an
eBPF program with a subset of the C language.


# Frequently asked questions

## Why is seccomp-bpf not enough?

A seccomp filter can access only raw syscall arguments (i.e. the
register values) which means that it is not possible to filter according
to the value pointed to by an argument, such as a file pathname. As an
embryonic Landlock version demonstrated, filtering at the syscall level
is complicated (e.g. need to take care of race conditions). This is
mainly because the access control checkpoints of the kernel are not at
this high-level but more underneath, at the LSM-hook level. The LSM
hooks are designed to handle this kind of checks.  Landlock abstracts
this approach to leverage the ability of unprivileged users to limit
themselves.

Cf. section "What it isn't?" in Documentation/prctl/seccomp_filter.txt


## Why use the seccomp(2) syscall?

Landlock use the same semantic as seccomp to apply access rule
restrictions. It add a new layer of security for the current process
which is inherited by its children. It makes sense to use an unique
access-restricting syscall (that should be allowed by seccomp filters)
which can only drop privileges. Moreover, a Landlock rule could come
from outside a process (e.g.  passed through a UNIX socket). It is then
useful to differentiate the creation/load of Landlock eBPF programs via
bpf(2), from rule enforcement via seccomp(2).


## Why a new LSM? Are SELinux, AppArmor, Smack and Tomoyo not good
   enough?

The current access control LSMs are fine for their purpose which is to
give the *root* the ability to enforce a security policy for the
*system*. What is missing is a way to enforce a security policy for any
application by its developer and *unprivileged user* as seccomp can do
for raw syscall filtering.

Differences from other (access control) LSMs:
* not only dedicated to administrators (i.e. no_new_priv);
* limited kernel attack surface (e.g. policy parsing);
* constrained policy rules (no DoS: deterministic execution time);
* do not leak more information than the loader process can legitimately
  have access to (minimize metadata inference).


# Changes since v9

* replace subtype with expected_attach_type and a new expected_attach_triggers
  and update libbpf accordingly
* handle inode and map lifetime with LSM hooks
* use a hash map for the inode map: integrate inodemap.c into hashtab.c
* allow arbitrary value size instead of 64-bits


# Changes since v8

* fit with the new LSM stacking framework (security blobs were tested
  but are not use in this series because of the code reduction)
* remove the Landlock program chaining and the file path evaluation
  feature to get a minimum viable product and ease the review
* replace the example with a simple blacklist policy
* rebase on bpf-next


# Changes since v7

* major revamp of the file system enforcement:
  * new eBPF map dedicated to tie an inode with an arbitrary 64-bits
    value, which can be used to tag files
  * three new Landlock hooks: FS_WALK, FS_PICK and FS_GET
  * add the ability to chain Landlock programs
  * add a new eBPF map type to compare inodes
  * don't use macros anymore
* replace subtype fields:
  * triggers: fine-grained bitfiel of actions on which a Landlock
    program may be called (if it comes from a sandbox process)
  * previous: a parent chained program
* upstreamed patches:
  * commit 369130b63178 ("selftests: Enhance kselftest_harness.h to
    print which assert failed")


# Changes since v6

* upstreamed patches:
  * commit 752ba56fb130 ("bpf: Extend check_uarg_tail_zero() checks")
  * commit 0b40808a1084 ("selftests: Make test_harness.h more generally
    available") and related ones
  * commit 3bb857e47e49 ("LSM: Enable multiple calls to
    security_add_hooks() for the same LSM")
* simplify the landlock_context (remove syscall_* fields) and add three
  FS sub-events: IOCTL, LOCK, FCNTL
* minimize the number of callable BPF functions from a Landlock rule
* do not split put_seccomp_filter() with put_seccomp()
* rename Landlock version to Landlock ABI
* miscellaneous fixes
* rebase on net-next


# Changes since v5

* eBPF program subtype:
  * use a prog_subtype pointer instead of inlining it into bpf_attr
  * enable a future-proof behavior (reject unhandled data/size)
  * add tests
* use a simple rule hierarchy (similar to seccomp-bpf)
* add a ptrace scope protection
* add more tests
* add more documentation
* rename some files
* miscellaneous fixes
* rebase on net-next


# Changes since v4

* upstreamed patches:
  * commit d498f8719a09 ("bpf: Rebuild bpf.o for any dependency update")
  * commit a734fb5d6006 ("samples/bpf: Reset global variables") and
    related ones
  * commit f4874d01beba ("bpf: Use bpf_create_map() from the library")
    and related ones
  * commit d02d8986a768 ("bpf: Always test unprivileged programs")
  * commit 640eb7e7b524 ("fs: Constify path_is_under()'s arguments")
  * commit 535e7b4b5ef2 ("bpf: Use u64_to_user_ptr()")
* revamp Landlock to not expose an LSM hook interface but wrap and
  abstract them with Landlock events (currently one for all filesystem
  related operations: LANDLOCK_SUBTYPE_EVENT_FS)
* wrap all filesystem kernel objects through the same FS handle (struct
  landlock_handle_fs): struct file, struct inode, struct path and struct
  dentry
* a rule don't return an errno code but only a boolean to allow or deny
  an access request
* handle all filesystem related LSM hooks
* add some tests and a sample:
  * BPF context tests
  * Landlock sandboxing tests and sample
  * write Landlock rules in C and compile them with LLVM
* change field names of eBPF program subtype
* remove arraymap of handles for now (will be replaced with a revamped
  map)
* remove cgroup handling for now
* add user and kernel documentation
* rebase on net-next


# Changes since v3

* upstreamed patch:
  * commit 1955351da41c ("bpf: Set register type according to
    is_valid_access()")
* use abstract LSM hook arguments with custom types (e.g.
  *_LANDLOCK_ARG_FS for struct file, struct inode and struct path)
* add more LSM hooks to support full filesystem access control
* improve the sandbox example
* fix races and RCU issues:
  * eBPF program execution and eBPF helpers
  * revamp the arraymap of handles to cleanly deal with update/delete
* eBPF program subtype for Landlock:
  * remove the "origin" field
  * add an "option" field
* rebase onto Daniel Mack's patches v7 [3]
* remove merged commit 1955351da41c ("bpf: Set register type according
  to is_valid_access()")
* fix spelling mistakes
* cleanup some type and variable names
* split patches
* for now, remove cgroup delegation handling for unprivileged user
* remove extra access check for cgroup_get_from_fd()
* remove unused example code dealing with skb
* remove seccomp-bpf link:
  * no more seccomp cookie
  * for now, it is no more possible to check the current syscall
    properties


# Changes since v2

* revamp cgroup handling:
  * use Daniel Mack's patches "Add eBPF hooks for cgroups" v5
  * remove bpf_landlock_cmp_cgroup_beneath()
  * make BPF_PROG_ATTACH usable with delegated cgroups
  * add a new CGRP_NO_NEW_PRIVS flag for safe cgroups
  * handle Landlock sandboxing for cgroups hierarchy
  * allow unprivileged processes to attach Landlock eBPF program to
    cgroups
* add subtype to eBPF programs:
  * replace Landlock hook identification by custom eBPF program types
    with a dedicated subtype field
  * manage fine-grained privileged Landlock programs
  * register Landlock programs for dedicated trigger origins (e.g.
    syscall, return from seccomp filter and/or interruption)
* performance and memory optimizations: use an array to access Landlock
  hooks directly but do not duplicated it for each thread
  (seccomp-based)
* allow running Landlock programs without seccomp filter
* fix seccomp-related issues
* remove extra errno bounding check for Landlock programs
* add some examples for optional eBPF functions or context access
  (network related) according to security checks to allow more features
  for privileged programs (e.g. Checmate)


# Changes since v1

* focus on the LSM hooks, not the syscalls:
  * much more simple implementation
  * does not need audit cache tricks to avoid race conditions
  * more simple to use and more generic because using the LSM hook
    abstraction directly
  * more efficient because only checking in LSM hooks
  * architecture agnostic
* switch from cBPF to eBPF:
  * new eBPF program types dedicated to Landlock
  * custom functions used by the eBPF program
  * gain some new features (e.g. 10 registers, can load values of
    different size, LLVM translator) but only a few functions allowed
    and a dedicated map type
  * new context: LSM hook ID, cookie and LSM hook arguments
  * need to set the sysctl kernel.unprivileged_bpf_disable to 0 (default
    value) to be able to load hook filters as unprivileged users
* smaller and simpler:
  * no more checker groups but dedicated arraymap of handles
  * simpler userland structs thanks to eBPF functions
* distinctive name: Landlock


[1] https://lore.kernel.org/linux-security-module/20190625215239.11136-1-mic@digikod.net/
[2] https://lore.kernel.org/lkml/5828776A.1010104@digikod.net/
[3] https://lore.kernel.org/netdev/1477390454-12553-1-git-send-email-daniel@zonque.org/
[4] https://lore.kernel.org/lkml/50db058a-7dde-441b-a7f9-f6837fe8b69f@schaufler-ca.com/
[5] https://lore.kernel.org/lkml/20190520133305.11925-1-cyphar@cyphar.com/

Regards,

Mickaël Salaün (10):
  fs,security: Add a new file access type: MAY_CHROOT
  bpf: Add expected_attach_triggers and a is_valid_triggers() verifier
  bpf,landlock: Define an eBPF program type for Landlock hooks
  seccomp,landlock: Enforce Landlock programs per process hierarchy
  landlock: Handle filesystem access control
  bpf,landlock: Add a new map type: inode
  landlock: Add ptrace restrictions
  bpf: Add a Landlock sandbox example
  bpf,landlock: Add tests for Landlock
  landlock: Add user and kernel documentation for Landlock

 Documentation/security/index.rst              |   1 +
 Documentation/security/landlock/index.rst     |  20 +
 Documentation/security/landlock/kernel.rst    |  99 +++
 Documentation/security/landlock/user.rst      | 147 ++++
 MAINTAINERS                                   |  13 +
 fs/open.c                                     |   3 +-
 include/linux/bpf.h                           |  18 +
 include/linux/bpf_types.h                     |   6 +
 include/linux/fs.h                            |   1 +
 include/linux/landlock.h                      |  38 ++
 include/linux/lsm_hooks.h                     |   1 +
 include/linux/seccomp.h                       |   5 +
 include/uapi/linux/bpf.h                      |  16 +-
 include/uapi/linux/landlock.h                 |  94 +++
 include/uapi/linux/seccomp.h                  |   1 +
 kernel/bpf/core.c                             |   2 +
 kernel/bpf/hashtab.c                          | 253 +++++++
 kernel/bpf/syscall.c                          |  41 +-
 kernel/bpf/verifier.c                         |  26 +
 kernel/fork.c                                 |   8 +-
 kernel/seccomp.c                              |   4 +
 samples/bpf/.gitignore                        |   1 +
 samples/bpf/Makefile                          |   3 +
 samples/bpf/landlock1.h                       |   8 +
 samples/bpf/landlock1_kern.c                  |  55 ++
 samples/bpf/landlock1_user.c                  | 250 +++++++
 security/Kconfig                              |   1 +
 security/Makefile                             |   2 +
 security/landlock/Kconfig                     |  18 +
 security/landlock/Makefile                    |   5 +
 security/landlock/common.h                    | 105 +++
 security/landlock/enforce.c                   | 272 ++++++++
 security/landlock/enforce.h                   |  18 +
 security/landlock/enforce_seccomp.c           |  92 +++
 security/landlock/hooks.c                     |  94 +++
 security/landlock/hooks.h                     |  31 +
 security/landlock/hooks_fs.c                  | 639 ++++++++++++++++++
 security/landlock/hooks_fs.h                  |  31 +
 security/landlock/hooks_ptrace.c              | 121 ++++
 security/landlock/hooks_ptrace.h              |   8 +
 security/landlock/init.c                      | 148 ++++
 security/security.c                           |  15 +
 tools/include/uapi/linux/bpf.h                |  16 +-
 tools/include/uapi/linux/landlock.h           | 109 +++
 tools/lib/bpf/bpf.h                           |   1 +
 tools/lib/bpf/libbpf.c                        |  44 +-
 tools/lib/bpf/libbpf.h                        |   7 +-
 tools/lib/bpf/libbpf.map                      |   2 +
 tools/lib/bpf/libbpf_probes.c                 |   2 +
 tools/testing/selftests/Makefile              |   1 +
 tools/testing/selftests/bpf/bpf_helpers.h     |   2 +
 .../selftests/bpf/test_section_names.c        |   2 +-
 .../selftests/bpf/test_sockopt_multi.c        |   4 +-
 tools/testing/selftests/bpf/test_sockopt_sk.c |   2 +-
 tools/testing/selftests/bpf/test_verifier.c   |   1 +
 .../testing/selftests/bpf/verifier/landlock.c |  24 +
 tools/testing/selftests/landlock/.gitignore   |   4 +
 tools/testing/selftests/landlock/Makefile     |  39 ++
 tools/testing/selftests/landlock/test.h       |  50 ++
 tools/testing/selftests/landlock/test_base.c  |  24 +
 tools/testing/selftests/landlock/test_fs.c    | 256 +++++++
 .../testing/selftests/landlock/test_ptrace.c  | 148 ++++
 62 files changed, 3432 insertions(+), 20 deletions(-)
 create mode 100644 Documentation/security/landlock/index.rst
 create mode 100644 Documentation/security/landlock/kernel.rst
 create mode 100644 Documentation/security/landlock/user.rst
 create mode 100644 include/linux/landlock.h
 create mode 100644 include/uapi/linux/landlock.h
 create mode 100644 samples/bpf/landlock1.h
 create mode 100644 samples/bpf/landlock1_kern.c
 create mode 100644 samples/bpf/landlock1_user.c
 create mode 100644 security/landlock/Kconfig
 create mode 100644 security/landlock/Makefile
 create mode 100644 security/landlock/common.h
 create mode 100644 security/landlock/enforce.c
 create mode 100644 security/landlock/enforce.h
 create mode 100644 security/landlock/enforce_seccomp.c
 create mode 100644 security/landlock/hooks.c
 create mode 100644 security/landlock/hooks.h
 create mode 100644 security/landlock/hooks_fs.c
 create mode 100644 security/landlock/hooks_fs.h
 create mode 100644 security/landlock/hooks_ptrace.c
 create mode 100644 security/landlock/hooks_ptrace.h
 create mode 100644 security/landlock/init.c
 create mode 100644 tools/include/uapi/linux/landlock.h
 create mode 100644 tools/testing/selftests/bpf/verifier/landlock.c
 create mode 100644 tools/testing/selftests/landlock/.gitignore
 create mode 100644 tools/testing/selftests/landlock/Makefile
 create mode 100644 tools/testing/selftests/landlock/test.h
 create mode 100644 tools/testing/selftests/landlock/test_base.c
 create mode 100644 tools/testing/selftests/landlock/test_fs.c
 create mode 100644 tools/testing/selftests/landlock/test_ptrace.c

-- 
2.22.0


^ permalink raw reply	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 01/10] fs,security: Add a new file access type: MAY_CHROOT
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 02/10] bpf: Add expected_attach_triggers and a is_valid_triggers() verifier Mickaël Salaün
                   ` (8 subsequent siblings)
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

For compatibility reason, MAY_CHROOT is always set with MAY_CHDIR.
However, this new flag enable to differentiate a chdir form a chroot.

This is needed for the Landlock LSM to be able to evaluate a new root
directory.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: James Morris <jmorris@namei.org>
Cc: John Johansen <john.johansen@canonical.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Paul Moore <paul@paul-moore.com>
Cc: "Serge E. Hallyn" <serge@hallyn.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: linux-fsdevel@vger.kernel.org
---
 fs/open.c          | 3 ++-
 include/linux/fs.h | 1 +
 2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/open.c b/fs/open.c
index b5b80469b93d..e8767318fd03 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -494,7 +494,8 @@ int ksys_chroot(const char __user *filename)
 	if (error)
 		goto out;
 
-	error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR);
+	error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR |
+			MAY_CHROOT);
 	if (error)
 		goto dput_and_out;
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 75f2ed289a3f..7a0d92b1da85 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -99,6 +99,7 @@ typedef int (dio_iodone_t)(struct kiocb *iocb, loff_t offset,
 #define MAY_CHDIR		0x00000040
 /* called from RCU mode, don't block */
 #define MAY_NOT_BLOCK		0x00000080
+#define MAY_CHROOT		0x00000100
 
 /*
  * flags in file.f_mode.  Note that FMODE_READ and FMODE_WRITE must correspond
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 02/10] bpf: Add expected_attach_triggers and a is_valid_triggers() verifier
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 01/10] fs,security: Add a new file access type: MAY_CHROOT Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 03/10] bpf,landlock: Define an eBPF program type for Landlock hooks Mickaël Salaün
                   ` (7 subsequent siblings)
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

The goal of the program triggers is to be able to have static triggers
(bitflags) conditionning an eBPF program interpretation.  This help to
avoid unnecessary runs.

The struct bpf_verifier_ops gets a new optional function:
is_valid_verifier(). This new verifier is called at the beginning of the
eBPF program verification to check if the (optional) program triggers
are valid.

For now, only Landlock eBPF programs are using program triggers (see
next commits) but this could be used by other program types in the
future.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Link: https://lkml.kernel.org/r/20160827205559.GA43880@ast-mbp.thefacebook.com
---

Changes since v9:
* replace subtype with expected_attach_type (suggested by Alexei
  Starovoitov) and a new expected_attach_triggers
* add new bpf_attach_type: BPF_LANDLOCK_FS_PICK and BPF_LANDLOCK_FS_WALK
* remove bpf_prog_extra from bpf_base_func_proto()
* update libbpf and test_verifier to handle triggers

Changes since v8:
* use bpf_load_program_xattr() instead of bpf_load_program() and add
  bpf_verify_program_xattr() to deal with subtypes
* remove put_extra() since there is no more "previous" field (for now)

Changes since v7:
* rename LANDLOCK_SUBTYPE_* to LANDLOCK_*
* move subtype in bpf_prog_aux and use only one bit for has_subtype
  (suggested by Alexei Starovoitov)
* wrap the prog_subtype with a prog_extra to be able to reference kernel
  pointers:
  * add an optional put_extra() function to struct bpf_prog_ops to be
    able to free the pointed data
  * replace all the prog_subtype with prog_extra in the struct
    bpf_verifier_ops functions
* remove the ABI field (requested by Alexei Starovoitov)
* rename subtype fields

Changes since v6:
* rename Landlock version to ABI to better reflect its purpose
* fix unsigned integer checks
* fix pointer cast
* constify pointers
* rebase

Changes since v5:
* use a prog_subtype pointer and make it future-proof
* add subtype test
* constify bpf_load_program()'s subtype argument
* cleanup subtype initialization
* rebase

Changes since v4:
* replace the "status" field with "version" (more generic)
* replace the "access" field with "ability" (less confusing)

Changes since v3:
* remove the "origin" field
* add an "option" field
* cleanup comments
---
 include/linux/bpf.h            |  2 ++
 include/linux/bpf_types.h      |  3 +++
 include/uapi/linux/bpf.h       |  3 +++
 kernel/bpf/syscall.c           | 14 +++++++++++++-
 kernel/bpf/verifier.c          | 11 +++++++++++
 tools/include/uapi/linux/bpf.h |  3 +++
 tools/lib/bpf/bpf.h            |  1 +
 tools/lib/bpf/libbpf.map       |  1 +
 8 files changed, 37 insertions(+), 1 deletion(-)

diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index 18f4cc2c6acd..6d9c7a08713e 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -319,6 +319,7 @@ struct bpf_verifier_ops {
 				  const struct bpf_insn *src,
 				  struct bpf_insn *dst,
 				  struct bpf_prog *prog, u32 *target_size);
+	bool (*is_valid_triggers)(const struct bpf_prog *prog);
 };
 
 struct bpf_prog_offload_ops {
@@ -418,6 +419,7 @@ struct bpf_prog_aux {
 		struct work_struct work;
 		struct rcu_head	rcu;
 	};
+	u64 expected_attach_triggers;
 };
 
 struct bpf_array {
diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h
index eec5aeeeaf92..2ab647323f3a 100644
--- a/include/linux/bpf_types.h
+++ b/include/linux/bpf_types.h
@@ -38,6 +38,9 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_LIRC_MODE2, lirc_mode2)
 #ifdef CONFIG_INET
 BPF_PROG_TYPE(BPF_PROG_TYPE_SK_REUSEPORT, sk_reuseport)
 #endif
+#ifdef CONFIG_SECURITY_LANDLOCK
+BPF_PROG_TYPE(BPF_PROG_TYPE_LANDLOCK_HOOK, landlock)
+#endif
 
 BPF_MAP_TYPE(BPF_MAP_TYPE_ARRAY, array_map_ops)
 BPF_MAP_TYPE(BPF_MAP_TYPE_PERCPU_ARRAY, percpu_array_map_ops)
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 6f68438aa4ed..1664d260861b 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -197,6 +197,8 @@ enum bpf_attach_type {
 	BPF_CGROUP_UDP6_RECVMSG,
 	BPF_CGROUP_GETSOCKOPT,
 	BPF_CGROUP_SETSOCKOPT,
+	BPF_LANDLOCK_FS_PICK,
+	BPF_LANDLOCK_FS_WALK,
 	__MAX_BPF_ATTACH_TYPE
 };
 
@@ -412,6 +414,7 @@ union bpf_attr {
 		__u32		line_info_rec_size;	/* userspace bpf_line_info size */
 		__aligned_u64	line_info;	/* line info */
 		__u32		line_info_cnt;	/* number of bpf_line_info records */
+		__aligned_u64	expected_attach_triggers;	/* bitfield of triggers, e.g. LANDLOCK_TRIGGER_* */
 	};
 
 	struct { /* anonymous struct used by BPF_OBJ_* commands */
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 5d141f16f6fa..b2a8cb14f28e 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -1598,13 +1598,23 @@ bpf_prog_load_check_attach_type(enum bpf_prog_type prog_type,
 		default:
 			return -EINVAL;
 		}
+#ifdef CONFIG_SECURITY_LANDLOCK
+	case BPF_PROG_TYPE_LANDLOCK_HOOK:
+		switch (expected_attach_type) {
+		case BPF_LANDLOCK_FS_PICK:
+		case BPF_LANDLOCK_FS_WALK:
+			return 0;
+		default:
+			return -EINVAL;
+		}
+#endif
 	default:
 		return 0;
 	}
 }
 
 /* last field in 'union bpf_attr' used by this command */
-#define	BPF_PROG_LOAD_LAST_FIELD line_info_cnt
+#define	BPF_PROG_LOAD_LAST_FIELD expected_attach_triggers
 
 static int bpf_prog_load(union bpf_attr *attr, union bpf_attr __user *uattr)
 {
@@ -1694,6 +1704,8 @@ static int bpf_prog_load(union bpf_attr *attr, union bpf_attr __user *uattr)
 	if (err)
 		goto free_prog;
 
+	prog->aux->expected_attach_triggers = attr->expected_attach_triggers;
+
 	/* run eBPF verifier */
 	err = bpf_check(&prog, attr, uattr);
 	if (err < 0)
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index a2e763703c30..94a43d7c8175 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -9265,6 +9265,17 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr,
 	if (ret < 0)
 		goto skip_full_check;
 
+	if (env->ops->is_valid_triggers) {
+		if (!env->ops->is_valid_triggers(env->prog)) {
+			ret = -EINVAL;
+			goto err_unlock;
+		}
+	} else if (env->prog->aux->expected_attach_triggers) {
+		/* do not accept triggers if they are not handled */
+		ret = -EINVAL;
+		goto err_unlock;
+	}
+
 	if (bpf_prog_is_dev_bound(env->prog->aux)) {
 		ret = bpf_prog_offload_verifier_prep(env->prog);
 		if (ret)
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index f506c68b2612..232747393405 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -197,6 +197,8 @@ enum bpf_attach_type {
 	BPF_CGROUP_UDP6_RECVMSG,
 	BPF_CGROUP_GETSOCKOPT,
 	BPF_CGROUP_SETSOCKOPT,
+	BPF_LANDLOCK_FS_PICK,
+	BPF_LANDLOCK_FS_WALK,
 	__MAX_BPF_ATTACH_TYPE
 };
 
@@ -412,6 +414,7 @@ union bpf_attr {
 		__u32		line_info_rec_size;	/* userspace bpf_line_info size */
 		__aligned_u64	line_info;	/* line info */
 		__u32		line_info_cnt;	/* number of bpf_line_info records */
+		__aligned_u64	expected_attach_triggers;	/* bitfield of triggers, e.g. LANDLOCK_TRIGGER_* */
 	};
 
 	struct { /* anonymous struct used by BPF_OBJ_* commands */
diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h
index ff42ca043dc8..468bb3ac0be0 100644
--- a/tools/lib/bpf/bpf.h
+++ b/tools/lib/bpf/bpf.h
@@ -102,6 +102,7 @@ LIBBPF_API int bpf_load_program(enum bpf_prog_type type,
 				const struct bpf_insn *insns, size_t insns_cnt,
 				const char *license, __u32 kern_version,
 				char *log_buf, size_t log_buf_sz);
+LIBBPF_API int bpf_verify_program_xattr(union bpf_attr *attr, size_t attr_sz);
 LIBBPF_API int bpf_verify_program(enum bpf_prog_type type,
 				  const struct bpf_insn *insns,
 				  size_t insns_cnt, __u32 prog_flags,
diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
index f9d316e873d8..36ac26bdfda0 100644
--- a/tools/lib/bpf/libbpf.map
+++ b/tools/lib/bpf/libbpf.map
@@ -107,6 +107,7 @@ LIBBPF_0.0.1 {
 		bpf_set_link_xdp_fd;
 		bpf_task_fd_query;
 		bpf_verify_program;
+		bpf_verify_program_xattr;
 		btf__fd;
 		btf__find_by_name;
 		btf__free;
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 03/10] bpf,landlock: Define an eBPF program type for Landlock hooks
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 01/10] fs,security: Add a new file access type: MAY_CHROOT Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 02/10] bpf: Add expected_attach_triggers and a is_valid_triggers() verifier Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 04/10] seccomp,landlock: Enforce Landlock programs per process hierarchy Mickaël Salaün
                   ` (6 subsequent siblings)
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

Add a new type of eBPF program used by Landlock hooks.  The goal of this
type of program is to accept or deny a requested access from userspace
to a kernel object (e.g. a file).

This new BPF program type will be registered with the Landlock LSM
initialization.

Add an initial Landlock Kconfig and update the MAINTAINERS file.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: James Morris <jmorris@namei.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
---

Changes since v9:
* handle inode put and map put, which fix unmount (reported by Al Viro)
* replace subtype with expected_attach_type and expected_attach_triggers
* check eBPF program return code

Changes since v8:
* Remove the chaining concept from the eBPF program contexts (chain and
  cookie). We need to keep these subtypes this way to be able to make
  them evolve, though.
* remove bpf_landlock_put_extra() because there is no more a "previous"
  field to free (for now)

Changes since v7:
* cosmetic fixes
* rename LANDLOCK_SUBTYPE_* to LANDLOCK_*
* cleanup UAPI definitions and move them from bpf.h to landlock.h
  (suggested by Alexei Starovoitov)
* disable Landlock by default (suggested by Alexei Starovoitov)
* rename BPF_PROG_TYPE_LANDLOCK_{RULE,HOOK}
* update the Kconfig
* update the MAINTAINERS file
* replace the IOCTL, LOCK and FCNTL events with FS_PICK, FS_WALK and
  FS_GET hook types
* add the ability to chain programs with an eBPF program file descriptor
  (i.e. the "previous" field in a Landlock subtype) and keep a state
  with a "cookie" value available from the context
* add a "triggers" subtype bitfield to match specific actions (e.g.
  append, chdir, read...)

Changes since v6:
* add 3 more sub-events: IOCTL, LOCK, FCNTL
  https://lkml.kernel.org/r/2fbc99a6-f190-f335-bd14-04bdeed35571@digikod.net
* rename LANDLOCK_VERSION to LANDLOCK_ABI to better reflect its purpose,
  and move it from landlock.h to common.h
* rename BPF_PROG_TYPE_LANDLOCK to BPF_PROG_TYPE_LANDLOCK_RULE: an eBPF
  program could be used for something else than a rule
* simplify struct landlock_context by removing the arch and syscall_nr fields
* remove all eBPF map functions call, remove ABILITY_WRITE
* refactor bpf_landlock_func_proto() (suggested by Kees Cook)
* constify pointers
* fix doc inclusion

Changes since v5:
* rename file hooks.c to init.c
* fix spelling

Changes since v4:
* merge a minimal (not enabled) LSM code and Kconfig in this commit

Changes since v3:
* split commit
* revamp the landlock_context:
  * add arch, syscall_nr and syscall_cmd (ioctl, fcntl…) to be able to
    cross-check action with the event type
  * replace args array with dedicated fields to ease the addition of new
    fields
---
 MAINTAINERS                         |  13 ++++
 include/uapi/linux/bpf.h            |   1 +
 include/uapi/linux/landlock.h       |  94 ++++++++++++++++++++++++
 kernel/bpf/verifier.c               |   1 +
 security/Kconfig                    |   1 +
 security/Makefile                   |   2 +
 security/landlock/Kconfig           |  18 +++++
 security/landlock/Makefile          |   3 +
 security/landlock/common.h          |  44 +++++++++++
 security/landlock/init.c            | 100 +++++++++++++++++++++++++
 tools/include/uapi/linux/bpf.h      |   1 +
 tools/include/uapi/linux/landlock.h | 109 ++++++++++++++++++++++++++++
 tools/lib/bpf/libbpf.c              |   1 +
 tools/lib/bpf/libbpf_probes.c       |   1 +
 14 files changed, 389 insertions(+)
 create mode 100644 include/uapi/linux/landlock.h
 create mode 100644 security/landlock/Kconfig
 create mode 100644 security/landlock/Makefile
 create mode 100644 security/landlock/common.h
 create mode 100644 security/landlock/init.c
 create mode 100644 tools/include/uapi/linux/landlock.h

diff --git a/MAINTAINERS b/MAINTAINERS
index 211ea3a199bd..d30b535a613a 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8947,6 +8947,19 @@ F:	net/core/skmsg.c
 F:	net/core/sock_map.c
 F:	net/ipv4/tcp_bpf.c
 
+LANDLOCK SECURITY MODULE
+M:	Mickaël Salaün <mic@digikod.net>
+S:	Supported
+F:	Documentation/security/landlock/
+F:	include/linux/landlock.h
+F:	include/uapi/linux/landlock.h
+F:	samples/bpf/landlock*
+F:	security/landlock/
+F:	tools/include/uapi/linux/landlock.h
+F:	tools/testing/selftests/landlock/
+K:	landlock
+K:	LANDLOCK
+
 LANTIQ / INTEL Ethernet drivers
 M:	Hauke Mehrtens <hauke@hauke-m.de>
 L:	netdev@vger.kernel.org
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 1664d260861b..d68613f737f3 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -171,6 +171,7 @@ enum bpf_prog_type {
 	BPF_PROG_TYPE_CGROUP_SYSCTL,
 	BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE,
 	BPF_PROG_TYPE_CGROUP_SOCKOPT,
+	BPF_PROG_TYPE_LANDLOCK_HOOK,
 };
 
 enum bpf_attach_type {
diff --git a/include/uapi/linux/landlock.h b/include/uapi/linux/landlock.h
new file mode 100644
index 000000000000..be3c4757a3ad
--- /dev/null
+++ b/include/uapi/linux/landlock.h
@@ -0,0 +1,94 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * Landlock - UAPI headers
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#ifndef _UAPI__LINUX_LANDLOCK_H__
+#define _UAPI__LINUX_LANDLOCK_H__
+
+#include <linux/types.h>
+
+#define LANDLOCK_RET_ALLOW	0
+#define LANDLOCK_RET_DENY	1
+
+/**
+ * DOC: landlock_triggers
+ *
+ * A landlock trigger is used as a bitmask in subtype.landlock_hook.triggers
+ * for a fs_pick program.  It defines a set of actions for which the program
+ * should verify an access request.
+ *
+ * - %LANDLOCK_TRIGGER_FS_PICK_APPEND
+ * - %LANDLOCK_TRIGGER_FS_PICK_CHDIR
+ * - %LANDLOCK_TRIGGER_FS_PICK_CHROOT
+ * - %LANDLOCK_TRIGGER_FS_PICK_CREATE
+ * - %LANDLOCK_TRIGGER_FS_PICK_EXECUTE
+ * - %LANDLOCK_TRIGGER_FS_PICK_FCNTL
+ * - %LANDLOCK_TRIGGER_FS_PICK_GETATTR
+ * - %LANDLOCK_TRIGGER_FS_PICK_IOCTL
+ * - %LANDLOCK_TRIGGER_FS_PICK_LINK
+ * - %LANDLOCK_TRIGGER_FS_PICK_LINKTO
+ * - %LANDLOCK_TRIGGER_FS_PICK_LOCK
+ * - %LANDLOCK_TRIGGER_FS_PICK_MAP
+ * - %LANDLOCK_TRIGGER_FS_PICK_MOUNTON
+ * - %LANDLOCK_TRIGGER_FS_PICK_OPEN
+ * - %LANDLOCK_TRIGGER_FS_PICK_READ
+ * - %LANDLOCK_TRIGGER_FS_PICK_READDIR
+ * - %LANDLOCK_TRIGGER_FS_PICK_RECEIVE
+ * - %LANDLOCK_TRIGGER_FS_PICK_RENAME
+ * - %LANDLOCK_TRIGGER_FS_PICK_RENAMETO
+ * - %LANDLOCK_TRIGGER_FS_PICK_RMDIR
+ * - %LANDLOCK_TRIGGER_FS_PICK_SETATTR
+ * - %LANDLOCK_TRIGGER_FS_PICK_TRANSFER
+ * - %LANDLOCK_TRIGGER_FS_PICK_UNLINK
+ * - %LANDLOCK_TRIGGER_FS_PICK_WRITE
+ */
+#define LANDLOCK_TRIGGER_FS_PICK_APPEND			(1ULL << 0)
+#define LANDLOCK_TRIGGER_FS_PICK_CHDIR			(1ULL << 1)
+#define LANDLOCK_TRIGGER_FS_PICK_CHROOT			(1ULL << 2)
+#define LANDLOCK_TRIGGER_FS_PICK_CREATE			(1ULL << 3)
+#define LANDLOCK_TRIGGER_FS_PICK_EXECUTE		(1ULL << 4)
+#define LANDLOCK_TRIGGER_FS_PICK_FCNTL			(1ULL << 5)
+#define LANDLOCK_TRIGGER_FS_PICK_GETATTR		(1ULL << 6)
+#define LANDLOCK_TRIGGER_FS_PICK_IOCTL			(1ULL << 7)
+#define LANDLOCK_TRIGGER_FS_PICK_LINK			(1ULL << 8)
+#define LANDLOCK_TRIGGER_FS_PICK_LINKTO			(1ULL << 9)
+#define LANDLOCK_TRIGGER_FS_PICK_LOCK			(1ULL << 10)
+#define LANDLOCK_TRIGGER_FS_PICK_MAP			(1ULL << 11)
+#define LANDLOCK_TRIGGER_FS_PICK_MOUNTON		(1ULL << 12)
+#define LANDLOCK_TRIGGER_FS_PICK_OPEN			(1ULL << 13)
+#define LANDLOCK_TRIGGER_FS_PICK_READ			(1ULL << 14)
+#define LANDLOCK_TRIGGER_FS_PICK_READDIR		(1ULL << 15)
+#define LANDLOCK_TRIGGER_FS_PICK_RECEIVE		(1ULL << 16)
+#define LANDLOCK_TRIGGER_FS_PICK_RENAME			(1ULL << 17)
+#define LANDLOCK_TRIGGER_FS_PICK_RENAMETO		(1ULL << 18)
+#define LANDLOCK_TRIGGER_FS_PICK_RMDIR			(1ULL << 19)
+#define LANDLOCK_TRIGGER_FS_PICK_SETATTR		(1ULL << 20)
+#define LANDLOCK_TRIGGER_FS_PICK_TRANSFER		(1ULL << 21)
+#define LANDLOCK_TRIGGER_FS_PICK_UNLINK			(1ULL << 22)
+#define LANDLOCK_TRIGGER_FS_PICK_WRITE			(1ULL << 23)
+
+/**
+ * struct landlock_ctx_fs_pick - context accessible to a fs_pick program
+ *
+ * @inode: pointer to the current kernel object that can be used to compare
+ *         inodes from an inode map.
+ */
+struct landlock_ctx_fs_pick {
+	__u64 inode;
+};
+
+/**
+ * struct landlock_ctx_fs_walk - context accessible to a fs_walk program
+ *
+ * @inode: pointer to the current kernel object that can be used to compare
+ *         inodes from an inode map.
+ */
+struct landlock_ctx_fs_walk {
+	__u64 inode;
+};
+
+#endif /* _UAPI__LINUX_LANDLOCK_H__ */
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 94a43d7c8175..026c68cb9116 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -6116,6 +6116,7 @@ static int check_return_code(struct bpf_verifier_env *env)
 	case BPF_PROG_TYPE_CGROUP_DEVICE:
 	case BPF_PROG_TYPE_CGROUP_SYSCTL:
 	case BPF_PROG_TYPE_CGROUP_SOCKOPT:
+	case BPF_PROG_TYPE_LANDLOCK_HOOK:
 		break;
 	default:
 		return 0;
diff --git a/security/Kconfig b/security/Kconfig
index 06a30851511a..549822ea60b7 100644
--- a/security/Kconfig
+++ b/security/Kconfig
@@ -237,6 +237,7 @@ source "security/apparmor/Kconfig"
 source "security/loadpin/Kconfig"
 source "security/yama/Kconfig"
 source "security/safesetid/Kconfig"
+source "security/landlock/Kconfig"
 
 source "security/integrity/Kconfig"
 
diff --git a/security/Makefile b/security/Makefile
index c598b904938f..396ff107f70d 100644
--- a/security/Makefile
+++ b/security/Makefile
@@ -11,6 +11,7 @@ subdir-$(CONFIG_SECURITY_APPARMOR)	+= apparmor
 subdir-$(CONFIG_SECURITY_YAMA)		+= yama
 subdir-$(CONFIG_SECURITY_LOADPIN)	+= loadpin
 subdir-$(CONFIG_SECURITY_SAFESETID)    += safesetid
+subdir-$(CONFIG_SECURITY_LANDLOCK)		+= landlock
 
 # always enable default capabilities
 obj-y					+= commoncap.o
@@ -27,6 +28,7 @@ obj-$(CONFIG_SECURITY_APPARMOR)		+= apparmor/
 obj-$(CONFIG_SECURITY_YAMA)		+= yama/
 obj-$(CONFIG_SECURITY_LOADPIN)		+= loadpin/
 obj-$(CONFIG_SECURITY_SAFESETID)       += safesetid/
+obj-$(CONFIG_SECURITY_LANDLOCK)	+= landlock/
 obj-$(CONFIG_CGROUP_DEVICE)		+= device_cgroup.o
 
 # Object integrity file lists
diff --git a/security/landlock/Kconfig b/security/landlock/Kconfig
new file mode 100644
index 000000000000..8bd103102008
--- /dev/null
+++ b/security/landlock/Kconfig
@@ -0,0 +1,18 @@
+config SECURITY_LANDLOCK
+	bool "Landlock support"
+	depends on SECURITY
+	depends on BPF_SYSCALL
+	depends on SECCOMP_FILTER
+	default n
+	help
+	  This selects Landlock, a programmatic access control.  It enables to
+	  restrict processes on the fly (i.e. create a sandbox).  The security
+	  policy is a set of eBPF programs, dedicated to deny a list of actions
+	  on specific kernel objects (e.g. file).
+
+	  You need to enable seccomp filter to apply a security policy to a
+	  process hierarchy (e.g. application with built-in sandboxing).
+
+	  See Documentation/security/landlock/ for further information.
+
+	  If you are unsure how to answer this question, answer N.
diff --git a/security/landlock/Makefile b/security/landlock/Makefile
new file mode 100644
index 000000000000..7205f9a7a2ee
--- /dev/null
+++ b/security/landlock/Makefile
@@ -0,0 +1,3 @@
+obj-$(CONFIG_SECURITY_LANDLOCK) := landlock.o
+
+landlock-y := init.o
diff --git a/security/landlock/common.h b/security/landlock/common.h
new file mode 100644
index 000000000000..80dc36f4d0ac
--- /dev/null
+++ b/security/landlock/common.h
@@ -0,0 +1,44 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Landlock LSM - private headers
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#ifndef _SECURITY_LANDLOCK_COMMON_H
+#define _SECURITY_LANDLOCK_COMMON_H
+
+#include <linux/bpf.h> /* enum bpf_attach_type */
+#include <linux/filter.h> /* bpf_prog */
+#include <linux/refcount.h> /* refcount_t */
+#include <uapi/linux/landlock.h> /* LANDLOCK_TRIGGER_* */
+
+#define LANDLOCK_NAME "landlock"
+
+/* UAPI bounds and bitmasks */
+
+#define _LANDLOCK_HOOK_LAST LANDLOCK_HOOK_FS_WALK
+
+#define _LANDLOCK_TRIGGER_FS_PICK_LAST	LANDLOCK_TRIGGER_FS_PICK_WRITE
+#define _LANDLOCK_TRIGGER_FS_PICK_MASK	((_LANDLOCK_TRIGGER_FS_PICK_LAST << 1ULL) - 1)
+
+enum landlock_hook_type {
+	LANDLOCK_HOOK_FS_PICK = 1,
+	LANDLOCK_HOOK_FS_WALK,
+};
+
+static inline enum landlock_hook_type get_hook_type(const struct bpf_prog *prog)
+{
+	switch (prog->expected_attach_type) {
+	case BPF_LANDLOCK_FS_PICK:
+		return LANDLOCK_HOOK_FS_PICK;
+	case BPF_LANDLOCK_FS_WALK:
+		return LANDLOCK_HOOK_FS_WALK;
+	default:
+		WARN_ON(1);
+		return BPF_LANDLOCK_FS_PICK;
+	}
+}
+
+#endif /* _SECURITY_LANDLOCK_COMMON_H */
diff --git a/security/landlock/init.c b/security/landlock/init.c
new file mode 100644
index 000000000000..8dfd5fea3c1f
--- /dev/null
+++ b/security/landlock/init.c
@@ -0,0 +1,100 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Landlock LSM - init
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#include <linux/bpf.h> /* enum bpf_access_type */
+#include <linux/capability.h> /* capable */
+#include <linux/filter.h> /* struct bpf_prog */
+
+#include "common.h" /* LANDLOCK_* */
+
+static bool bpf_landlock_is_valid_access(int off, int size,
+		enum bpf_access_type type, const struct bpf_prog *prog,
+		struct bpf_insn_access_aux *info)
+{
+	enum bpf_reg_type reg_type = NOT_INIT;
+	int max_size = 0;
+
+	if (WARN_ON(!prog->expected_attach_type))
+		return false;
+
+	if (off < 0)
+		return false;
+	if (size <= 0 || size > sizeof(__u64))
+		return false;
+
+	/* check memory range access */
+	switch (reg_type) {
+	case NOT_INIT:
+		return false;
+	case SCALAR_VALUE:
+		/* allow partial raw value */
+		if (size > max_size)
+			return false;
+		info->ctx_field_size = max_size;
+		break;
+	default:
+		/* deny partial pointer */
+		if (size != max_size)
+			return false;
+	}
+
+	info->reg_type = reg_type;
+	return true;
+}
+
+static bool bpf_landlock_is_valid_triggers(const struct bpf_prog *prog)
+{
+	u64 triggers;
+
+	if (!prog)
+		return false;
+	triggers = prog->aux->expected_attach_triggers;
+
+	switch (get_hook_type(prog)) {
+	case LANDLOCK_HOOK_FS_PICK:
+		if (!triggers || triggers & ~_LANDLOCK_TRIGGER_FS_PICK_MASK)
+			return false;
+		break;
+	case LANDLOCK_HOOK_FS_WALK:
+		if (triggers)
+			return false;
+		break;
+	}
+	return true;
+}
+
+static const struct bpf_func_proto *bpf_landlock_func_proto(
+		enum bpf_func_id func_id,
+		const struct bpf_prog *prog)
+{
+	if (WARN_ON(!prog->expected_attach_type))
+		return NULL;
+
+	/* generic functions */
+	/* TODO: do we need/want update/delete functions for every LL prog?
+	 * => impurity vs. audit */
+	switch (func_id) {
+	case BPF_FUNC_map_lookup_elem:
+		return &bpf_map_lookup_elem_proto;
+	case BPF_FUNC_map_update_elem:
+		return &bpf_map_update_elem_proto;
+	case BPF_FUNC_map_delete_elem:
+		return &bpf_map_delete_elem_proto;
+	default:
+		break;
+	}
+	return NULL;
+}
+
+const struct bpf_verifier_ops landlock_verifier_ops = {
+	.get_func_proto	= bpf_landlock_func_proto,
+	.is_valid_access = bpf_landlock_is_valid_access,
+	.is_valid_triggers = bpf_landlock_is_valid_triggers,
+};
+
+const struct bpf_prog_ops landlock_prog_ops = {};
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index 232747393405..7b7a4f6c3104 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -171,6 +171,7 @@ enum bpf_prog_type {
 	BPF_PROG_TYPE_CGROUP_SYSCTL,
 	BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE,
 	BPF_PROG_TYPE_CGROUP_SOCKOPT,
+	BPF_PROG_TYPE_LANDLOCK_HOOK,
 };
 
 enum bpf_attach_type {
diff --git a/tools/include/uapi/linux/landlock.h b/tools/include/uapi/linux/landlock.h
new file mode 100644
index 000000000000..9e6d8e10ec2c
--- /dev/null
+++ b/tools/include/uapi/linux/landlock.h
@@ -0,0 +1,109 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * Landlock - UAPI headers
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#ifndef _UAPI__LINUX_LANDLOCK_H__
+#define _UAPI__LINUX_LANDLOCK_H__
+
+#include <linux/types.h>
+
+#define LANDLOCK_RET_ALLOW	0
+#define LANDLOCK_RET_DENY	1
+
+/**
+ * enum landlock_hook_type - hook type for which a Landlock program is called
+ *
+ * A hook is a policy decision point which exposes the same context type for
+ * each program evaluation.
+ *
+ * @LANDLOCK_HOOK_FS_PICK: called for the last element of a file path
+ * @LANDLOCK_HOOK_FS_WALK: called for each directory of a file path (excluding
+ *			   the directory passed to fs_pick, if any)
+ */
+enum landlock_hook_type {
+	LANDLOCK_HOOK_FS_PICK = 1,
+	LANDLOCK_HOOK_FS_WALK,
+};
+
+/**
+ * DOC: landlock_triggers
+ *
+ * A landlock trigger is used as a bitmask in subtype.landlock_hook.triggers
+ * for a fs_pick program.  It defines a set of actions for which the program
+ * should verify an access request.
+ *
+ * - %LANDLOCK_TRIGGER_FS_PICK_APPEND
+ * - %LANDLOCK_TRIGGER_FS_PICK_CHDIR
+ * - %LANDLOCK_TRIGGER_FS_PICK_CHROOT
+ * - %LANDLOCK_TRIGGER_FS_PICK_CREATE
+ * - %LANDLOCK_TRIGGER_FS_PICK_EXECUTE
+ * - %LANDLOCK_TRIGGER_FS_PICK_FCNTL
+ * - %LANDLOCK_TRIGGER_FS_PICK_GETATTR
+ * - %LANDLOCK_TRIGGER_FS_PICK_IOCTL
+ * - %LANDLOCK_TRIGGER_FS_PICK_LINK
+ * - %LANDLOCK_TRIGGER_FS_PICK_LINKTO
+ * - %LANDLOCK_TRIGGER_FS_PICK_LOCK
+ * - %LANDLOCK_TRIGGER_FS_PICK_MAP
+ * - %LANDLOCK_TRIGGER_FS_PICK_MOUNTON
+ * - %LANDLOCK_TRIGGER_FS_PICK_OPEN
+ * - %LANDLOCK_TRIGGER_FS_PICK_READ
+ * - %LANDLOCK_TRIGGER_FS_PICK_READDIR
+ * - %LANDLOCK_TRIGGER_FS_PICK_RECEIVE
+ * - %LANDLOCK_TRIGGER_FS_PICK_RENAME
+ * - %LANDLOCK_TRIGGER_FS_PICK_RENAMETO
+ * - %LANDLOCK_TRIGGER_FS_PICK_RMDIR
+ * - %LANDLOCK_TRIGGER_FS_PICK_SETATTR
+ * - %LANDLOCK_TRIGGER_FS_PICK_TRANSFER
+ * - %LANDLOCK_TRIGGER_FS_PICK_UNLINK
+ * - %LANDLOCK_TRIGGER_FS_PICK_WRITE
+ */
+#define LANDLOCK_TRIGGER_FS_PICK_APPEND			(1ULL << 0)
+#define LANDLOCK_TRIGGER_FS_PICK_CHDIR			(1ULL << 1)
+#define LANDLOCK_TRIGGER_FS_PICK_CHROOT			(1ULL << 2)
+#define LANDLOCK_TRIGGER_FS_PICK_CREATE			(1ULL << 3)
+#define LANDLOCK_TRIGGER_FS_PICK_EXECUTE		(1ULL << 4)
+#define LANDLOCK_TRIGGER_FS_PICK_FCNTL			(1ULL << 5)
+#define LANDLOCK_TRIGGER_FS_PICK_GETATTR		(1ULL << 6)
+#define LANDLOCK_TRIGGER_FS_PICK_IOCTL			(1ULL << 7)
+#define LANDLOCK_TRIGGER_FS_PICK_LINK			(1ULL << 8)
+#define LANDLOCK_TRIGGER_FS_PICK_LINKTO			(1ULL << 9)
+#define LANDLOCK_TRIGGER_FS_PICK_LOCK			(1ULL << 10)
+#define LANDLOCK_TRIGGER_FS_PICK_MAP			(1ULL << 11)
+#define LANDLOCK_TRIGGER_FS_PICK_MOUNTON		(1ULL << 12)
+#define LANDLOCK_TRIGGER_FS_PICK_OPEN			(1ULL << 13)
+#define LANDLOCK_TRIGGER_FS_PICK_READ			(1ULL << 14)
+#define LANDLOCK_TRIGGER_FS_PICK_READDIR		(1ULL << 15)
+#define LANDLOCK_TRIGGER_FS_PICK_RECEIVE		(1ULL << 16)
+#define LANDLOCK_TRIGGER_FS_PICK_RENAME			(1ULL << 17)
+#define LANDLOCK_TRIGGER_FS_PICK_RENAMETO		(1ULL << 18)
+#define LANDLOCK_TRIGGER_FS_PICK_RMDIR			(1ULL << 19)
+#define LANDLOCK_TRIGGER_FS_PICK_SETATTR		(1ULL << 20)
+#define LANDLOCK_TRIGGER_FS_PICK_TRANSFER		(1ULL << 21)
+#define LANDLOCK_TRIGGER_FS_PICK_UNLINK			(1ULL << 22)
+#define LANDLOCK_TRIGGER_FS_PICK_WRITE			(1ULL << 23)
+
+/**
+ * struct landlock_ctx_fs_pick - context accessible to a fs_pick program
+ *
+ * @inode: pointer to the current kernel object that can be used to compare
+ *         inodes from an inode map.
+ */
+struct landlock_ctx_fs_pick {
+	__u64 inode;
+};
+
+/**
+ * struct landlock_ctx_fs_walk - context accessible to a fs_walk program
+ *
+ * @inode: pointer to the current kernel object that can be used to compare
+ *         inodes from an inode map.
+ */
+struct landlock_ctx_fs_walk {
+	__u64 inode;
+};
+
+#endif /* _UAPI__LINUX_LANDLOCK_H__ */
diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index ed07789b3e62..ab3b8b510b8a 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -2665,6 +2665,7 @@ static bool bpf_prog_type__needs_kver(enum bpf_prog_type type)
 	case BPF_PROG_TYPE_PERF_EVENT:
 	case BPF_PROG_TYPE_CGROUP_SYSCTL:
 	case BPF_PROG_TYPE_CGROUP_SOCKOPT:
+	case BPF_PROG_TYPE_LANDLOCK_HOOK:
 		return false;
 	case BPF_PROG_TYPE_KPROBE:
 	default:
diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c
index ace1a0708d99..03c910d1f84c 100644
--- a/tools/lib/bpf/libbpf_probes.c
+++ b/tools/lib/bpf/libbpf_probes.c
@@ -102,6 +102,7 @@ probe_load(enum bpf_prog_type prog_type, const struct bpf_insn *insns,
 	case BPF_PROG_TYPE_FLOW_DISSECTOR:
 	case BPF_PROG_TYPE_CGROUP_SYSCTL:
 	case BPF_PROG_TYPE_CGROUP_SOCKOPT:
+	case BPF_PROG_TYPE_LANDLOCK_HOOK:
 	default:
 		break;
 	}
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 04/10] seccomp,landlock: Enforce Landlock programs per process hierarchy
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
                   ` (2 preceding siblings ...)
  2019-07-21 21:31 ` [PATCH bpf-next v10 03/10] bpf,landlock: Define an eBPF program type for Landlock hooks Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 05/10] landlock: Handle filesystem access control Mickaël Salaün
                   ` (5 subsequent siblings)
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

The seccomp(2) syscall can be used by a task to apply a Landlock program
to itself. As a seccomp filter, a Landlock program is enforced for the
current task and all its future children. A program is immutable and a
task can only add new restricting programs to itself, forming a list of
programss.

A Landlock program is tied to a Landlock hook. If the action on a kernel
object is allowed by the other Linux security mechanisms (e.g. DAC,
capabilities, other LSM), then a Landlock hook related to this kind of
object is triggered. The list of programs for this hook is then
evaluated. Each program return a binary value which can deny the action
on a kernel object with a non-zero value. If every programs of the list
return zero, then the action on the object is allowed.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: James Morris <jmorris@namei.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
Cc: Will Drewry <wad@chromium.org>
Link: https://lkml.kernel.org/r/c10a503d-5e35-7785-2f3d-25ed8dd63fab@digikod.net
---

Changes since v9:
* replace subtype with expected_attach_type and expected_attach_triggers

Changes since v8:
* Remove the chaining concept from the eBPF program contexts (chain and
  cookie). We need to keep these subtypes this way to be able to make
  them evolve, though.

Changes since v7:
* handle and verify program chains
* split and rename providers.c to enforce.c and enforce_seccomp.c
* rename LANDLOCK_SUBTYPE_* to LANDLOCK_*

Changes since v6:
* rename some functions with more accurate names to reflect that an eBPF
  program for Landlock could be used for something else than a rule
* reword rule "appending" to "prepending" and explain it
* remove the superfluous no_new_privs check, only check global
  CAP_SYS_ADMIN when prepending a Landlock rule (needed for containers)
* create and use {get,put}_seccomp_landlock() (suggested by Kees Cook)
* replace ifdef with static inlined function (suggested by Kees Cook)
* use get_user() (suggested by Kees Cook)
* replace atomic_t with refcount_t (requested by Kees Cook)
* move struct landlock_{rule,events} from landlock.h to common.h
* cleanup headers

Changes since v5:
* remove struct landlock_node and use a similar inheritance mechanisme
  as seccomp-bpf (requested by Andy Lutomirski)
* rename SECCOMP_ADD_LANDLOCK_RULE to SECCOMP_APPEND_LANDLOCK_RULE
* rename file manager.c to providers.c
* add comments
* typo and cosmetic fixes

Changes since v4:
* merge manager and seccomp patches
* return -EFAULT in seccomp(2) when user_bpf_fd is null to easely check
  if Landlock is supported
* only allow a process with the global CAP_SYS_ADMIN to use Landlock
  (will be lifted in the future)
* add an early check to exit as soon as possible if the current process
  does not have Landlock rules

Changes since v3:
* remove the hard link with seccomp (suggested by Andy Lutomirski and
  Kees Cook):
  * remove the cookie which could imply multiple evaluation of Landlock
    rules
  * remove the origin field in struct landlock_data
* remove documentation fix (merged upstream)
* rename the new seccomp command to SECCOMP_ADD_LANDLOCK_RULE
* internal renaming
* split commit
* new design to be able to inherit on the fly the parent rules

Changes since v2:
* Landlock programs can now be run without seccomp filter but for any
  syscall (from the process) or interruption
* move Landlock related functions and structs into security/landlock/*
  (to manage cgroups as well)
* fix seccomp filter handling: run Landlock programs for each of their
  legitimate seccomp filter
* properly clean up all seccomp results
* cosmetic changes to ease the understanding
* fix some ifdef
---
 include/linux/landlock.h            |  34 ++++
 include/linux/seccomp.h             |   5 +
 include/uapi/linux/seccomp.h        |   1 +
 kernel/fork.c                       |   8 +-
 kernel/seccomp.c                    |   4 +
 security/landlock/Makefile          |   3 +-
 security/landlock/common.h          |  38 ++++
 security/landlock/enforce.c         | 272 ++++++++++++++++++++++++++++
 security/landlock/enforce.h         |  18 ++
 security/landlock/enforce_seccomp.c |  92 ++++++++++
 10 files changed, 473 insertions(+), 2 deletions(-)
 create mode 100644 include/linux/landlock.h
 create mode 100644 security/landlock/enforce.c
 create mode 100644 security/landlock/enforce.h
 create mode 100644 security/landlock/enforce_seccomp.c

diff --git a/include/linux/landlock.h b/include/linux/landlock.h
new file mode 100644
index 000000000000..8ac7942f50fc
--- /dev/null
+++ b/include/linux/landlock.h
@@ -0,0 +1,34 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Landlock LSM - public kernel headers
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#ifndef _LINUX_LANDLOCK_H
+#define _LINUX_LANDLOCK_H
+
+#include <linux/errno.h>
+#include <linux/sched.h> /* task_struct */
+
+#if defined(CONFIG_SECCOMP_FILTER) && defined(CONFIG_SECURITY_LANDLOCK)
+extern int landlock_seccomp_prepend_prog(unsigned int flags,
+		const int __user *user_bpf_fd);
+extern void put_seccomp_landlock(struct task_struct *tsk);
+extern void get_seccomp_landlock(struct task_struct *tsk);
+#else /* CONFIG_SECCOMP_FILTER && CONFIG_SECURITY_LANDLOCK */
+static inline int landlock_seccomp_prepend_prog(unsigned int flags,
+		const int __user *user_bpf_fd)
+{
+		return -EINVAL;
+}
+static inline void put_seccomp_landlock(struct task_struct *tsk)
+{
+}
+static inline void get_seccomp_landlock(struct task_struct *tsk)
+{
+}
+#endif /* CONFIG_SECCOMP_FILTER && CONFIG_SECURITY_LANDLOCK */
+
+#endif /* _LINUX_LANDLOCK_H */
diff --git a/include/linux/seccomp.h b/include/linux/seccomp.h
index 84868d37b35d..106a0ceff3d7 100644
--- a/include/linux/seccomp.h
+++ b/include/linux/seccomp.h
@@ -11,6 +11,7 @@
 
 #ifdef CONFIG_SECCOMP
 
+#include <linux/landlock.h>
 #include <linux/thread_info.h>
 #include <asm/seccomp.h>
 
@@ -22,6 +23,7 @@ struct seccomp_filter;
  *         system calls available to a process.
  * @filter: must always point to a valid seccomp-filter or NULL as it is
  *          accessed without locking during system call entry.
+ * @landlock_prog_set: contains a set of Landlock programs.
  *
  *          @filter must only be accessed from the context of current as there
  *          is no read locking.
@@ -29,6 +31,9 @@ struct seccomp_filter;
 struct seccomp {
 	int mode;
 	struct seccomp_filter *filter;
+#if defined(CONFIG_SECCOMP_FILTER) && defined(CONFIG_SECURITY_LANDLOCK)
+	struct landlock_prog_set *landlock_prog_set;
+#endif /* CONFIG_SECCOMP_FILTER && CONFIG_SECURITY_LANDLOCK */
 };
 
 #ifdef CONFIG_HAVE_ARCH_SECCOMP_FILTER
diff --git a/include/uapi/linux/seccomp.h b/include/uapi/linux/seccomp.h
index 90734aa5aa36..bce6534e7feb 100644
--- a/include/uapi/linux/seccomp.h
+++ b/include/uapi/linux/seccomp.h
@@ -16,6 +16,7 @@
 #define SECCOMP_SET_MODE_FILTER		1
 #define SECCOMP_GET_ACTION_AVAIL	2
 #define SECCOMP_GET_NOTIF_SIZES		3
+#define SECCOMP_PREPEND_LANDLOCK_PROG	4
 
 /* Valid flags for SECCOMP_SET_MODE_FILTER */
 #define SECCOMP_FILTER_FLAG_TSYNC		(1UL << 0)
diff --git a/kernel/fork.c b/kernel/fork.c
index 8f3e2d97d771..6c43517abdb9 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -51,6 +51,7 @@
 #include <linux/security.h>
 #include <linux/hugetlb.h>
 #include <linux/seccomp.h>
+#include <linux/landlock.h>
 #include <linux/swap.h>
 #include <linux/syscalls.h>
 #include <linux/jiffies.h>
@@ -458,6 +459,7 @@ void free_task(struct task_struct *tsk)
 	rt_mutex_debug_task_free(tsk);
 	ftrace_graph_exit_task(tsk);
 	put_seccomp_filter(tsk);
+	put_seccomp_landlock(tsk);
 	arch_release_task_struct(tsk);
 	if (tsk->flags & PF_KTHREAD)
 		free_kthread_struct(tsk);
@@ -888,7 +890,10 @@ static struct task_struct *dup_task_struct(struct task_struct *orig, int node)
 	 * the usage counts on the error path calling free_task.
 	 */
 	tsk->seccomp.filter = NULL;
-#endif
+#ifdef CONFIG_SECURITY_LANDLOCK
+	tsk->seccomp.landlock_prog_set = NULL;
+#endif /* CONFIG_SECURITY_LANDLOCK */
+#endif /* CONFIG_SECCOMP */
 
 	setup_thread_stack(tsk, orig);
 	clear_user_return_notifier(tsk);
@@ -1604,6 +1609,7 @@ static void copy_seccomp(struct task_struct *p)
 
 	/* Ref-count the new filter user, and assign it. */
 	get_seccomp_filter(current);
+	get_seccomp_landlock(current);
 	p->seccomp = current->seccomp;
 
 	/*
diff --git a/kernel/seccomp.c b/kernel/seccomp.c
index dba52a7db5e8..af542a2d21e7 100644
--- a/kernel/seccomp.c
+++ b/kernel/seccomp.c
@@ -41,6 +41,7 @@
 #include <linux/tracehook.h>
 #include <linux/uaccess.h>
 #include <linux/anon_inodes.h>
+#include <linux/landlock.h>
 
 enum notify_state {
 	SECCOMP_NOTIFY_INIT,
@@ -1397,6 +1398,9 @@ static long do_seccomp(unsigned int op, unsigned int flags,
 			return -EINVAL;
 
 		return seccomp_get_notif_sizes(uargs);
+	case SECCOMP_PREPEND_LANDLOCK_PROG:
+		return landlock_seccomp_prepend_prog(flags,
+				(const int __user *)uargs);
 	default:
 		return -EINVAL;
 	}
diff --git a/security/landlock/Makefile b/security/landlock/Makefile
index 7205f9a7a2ee..2a1a7082a365 100644
--- a/security/landlock/Makefile
+++ b/security/landlock/Makefile
@@ -1,3 +1,4 @@
 obj-$(CONFIG_SECURITY_LANDLOCK) := landlock.o
 
-landlock-y := init.o
+landlock-y := init.o \
+	enforce.o enforce_seccomp.o
diff --git a/security/landlock/common.h b/security/landlock/common.h
index 80dc36f4d0ac..2cf36dbf4560 100644
--- a/security/landlock/common.h
+++ b/security/landlock/common.h
@@ -28,6 +28,44 @@ enum landlock_hook_type {
 	LANDLOCK_HOOK_FS_WALK,
 };
 
+struct landlock_prog_list {
+	struct landlock_prog_list *prev;
+	struct bpf_prog *prog;
+	refcount_t usage;
+};
+
+/**
+ * struct landlock_prog_set - Landlock programs enforced on a thread
+ *
+ * This is used for low performance impact when forking a process. Instead of
+ * copying the full array and incrementing the usage of each entries, only
+ * create a pointer to &struct landlock_prog_set and increments its usage. When
+ * prepending a new program, if &struct landlock_prog_set is shared with other
+ * tasks, then duplicate it and prepend the program to this new &struct
+ * landlock_prog_set.
+ *
+ * @usage: reference count to manage the object lifetime. When a thread need to
+ *	   add Landlock programs and if @usage is greater than 1, then the
+ *	   thread must duplicate &struct landlock_prog_set to not change the
+ *	   children's programs as well.
+ * @programs: array of non-NULL &struct landlock_prog_list pointers
+ */
+struct landlock_prog_set {
+	struct landlock_prog_list *programs[_LANDLOCK_HOOK_LAST];
+	refcount_t usage;
+};
+
+/**
+ * get_hook_index - get an index for the programs of struct landlock_prog_set
+ *
+ * @type: a Landlock hook type
+ */
+static inline int get_hook_index(enum landlock_hook_type type)
+{
+	/* type ID > 0 for loaded programs */
+	return type - 1;
+}
+
 static inline enum landlock_hook_type get_hook_type(const struct bpf_prog *prog)
 {
 	switch (prog->expected_attach_type) {
diff --git a/security/landlock/enforce.c b/security/landlock/enforce.c
new file mode 100644
index 000000000000..b6979de69d3e
--- /dev/null
+++ b/security/landlock/enforce.c
@@ -0,0 +1,272 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Landlock LSM - enforcing helpers
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#include <asm/barrier.h> /* smp_store_release() */
+#include <asm/page.h> /* PAGE_SIZE */
+#include <linux/bpf.h> /* bpf_prog_put() */
+#include <linux/compiler.h> /* READ_ONCE() */
+#include <linux/err.h> /* PTR_ERR() */
+#include <linux/errno.h>
+#include <linux/filter.h> /* struct bpf_prog */
+#include <linux/refcount.h>
+#include <linux/slab.h> /* alloc(), kfree() */
+
+#include "common.h" /* struct landlock_prog_list */
+
+/* TODO: use a dedicated kmem_cache_alloc() instead of k*alloc() */
+
+static void put_landlock_prog_list(struct landlock_prog_list *prog_list)
+{
+	struct landlock_prog_list *orig = prog_list;
+
+	/* clean up single-reference branches iteratively */
+	while (orig && refcount_dec_and_test(&orig->usage)) {
+		struct landlock_prog_list *freeme = orig;
+
+		if (orig->prog)
+			bpf_prog_put(orig->prog);
+		orig = orig->prev;
+		kfree(freeme);
+	}
+}
+
+void landlock_put_prog_set(struct landlock_prog_set *prog_set)
+{
+	if (prog_set && refcount_dec_and_test(&prog_set->usage)) {
+		size_t i;
+
+		for (i = 0; i < ARRAY_SIZE(prog_set->programs); i++)
+			put_landlock_prog_list(prog_set->programs[i]);
+		kfree(prog_set);
+	}
+}
+
+void landlock_get_prog_set(struct landlock_prog_set *prog_set)
+{
+	if (!prog_set)
+		return;
+	refcount_inc(&prog_set->usage);
+}
+
+static struct landlock_prog_set *new_landlock_prog_set(void)
+{
+	struct landlock_prog_set *ret;
+
+	/* array filled with NULL values */
+	ret = kzalloc(sizeof(*ret), GFP_KERNEL);
+	if (!ret)
+		return ERR_PTR(-ENOMEM);
+	refcount_set(&ret->usage, 1);
+	return ret;
+}
+
+/**
+ * store_landlock_prog - prepend and deduplicate a Landlock prog_list
+ *
+ * Prepend @prog to @init_prog_set while ignoring @prog
+ * if they are already in @ref_prog_set.  Whatever is the result of this
+ * function call, you can call bpf_prog_put(@prog) after.
+ *
+ * @init_prog_set: empty prog_set to prepend to
+ * @ref_prog_set: prog_set to check for duplicate programs
+ * @prog: program to prepend
+ *
+ * Return -errno on error or 0 if @prog was successfully stored.
+ */
+static int store_landlock_prog(struct landlock_prog_set *init_prog_set,
+		const struct landlock_prog_set *ref_prog_set,
+		struct bpf_prog *prog)
+{
+	struct landlock_prog_list *tmp_list = NULL;
+	int err;
+	u32 hook_idx;
+	enum landlock_hook_type last_type;
+	struct bpf_prog *new = prog;
+
+	/* allocate all the memory we need */
+	struct landlock_prog_list *new_list;
+
+	last_type = get_hook_type(new);
+
+	/* ignore duplicate programs */
+	if (ref_prog_set) {
+		struct landlock_prog_list *ref;
+
+		hook_idx = get_hook_index(get_hook_type(new));
+		for (ref = ref_prog_set->programs[hook_idx];
+				ref; ref = ref->prev) {
+			if (ref->prog == new)
+				return -EINVAL;
+		}
+	}
+
+	new = bpf_prog_inc(new);
+	if (IS_ERR(new)) {
+		err = PTR_ERR(new);
+		goto put_tmp_list;
+	}
+	new_list = kzalloc(sizeof(*new_list), GFP_KERNEL);
+	if (!new_list) {
+		bpf_prog_put(new);
+		err = -ENOMEM;
+		goto put_tmp_list;
+	}
+	/* ignore Landlock types in this tmp_list */
+	new_list->prog = new;
+	new_list->prev = tmp_list;
+	refcount_set(&new_list->usage, 1);
+	tmp_list = new_list;
+
+	if (!tmp_list)
+		/* inform user space that this program was already added */
+		return -EEXIST;
+
+	/* properly store the list (without error cases) */
+	while (tmp_list) {
+		struct landlock_prog_list *new_list;
+
+		new_list = tmp_list;
+		tmp_list = tmp_list->prev;
+		/* do not increment the previous prog list usage */
+		hook_idx = get_hook_index(get_hook_type(new_list->prog));
+		new_list->prev = init_prog_set->programs[hook_idx];
+		/* no need to add from the last program to the first because
+		 * each of them are a different Landlock type */
+		smp_store_release(&init_prog_set->programs[hook_idx], new_list);
+	}
+	return 0;
+
+put_tmp_list:
+	put_landlock_prog_list(tmp_list);
+	return err;
+}
+
+/* limit Landlock programs set to 256KB */
+#define LANDLOCK_PROGRAMS_MAX_PAGES (1 << 6)
+
+/**
+ * landlock_prepend_prog - attach a Landlock prog_list to @current_prog_set
+ *
+ * Whatever is the result of this function call, you can call
+ * bpf_prog_put(@prog) after.
+ *
+ * @current_prog_set: landlock_prog_set pointer, must be locked (if needed) to
+ *                    prevent a concurrent put/free. This pointer must not be
+ *                    freed after the call.
+ * @prog: non-NULL Landlock prog_list to prepend to @current_prog_set. @prog
+ *	  will be owned by landlock_prepend_prog() and freed if an error
+ *	  happened.
+ *
+ * Return @current_prog_set or a new pointer when OK. Return a pointer error
+ * otherwise.
+ */
+struct landlock_prog_set *landlock_prepend_prog(
+		struct landlock_prog_set *current_prog_set,
+		struct bpf_prog *prog)
+{
+	struct landlock_prog_set *new_prog_set = current_prog_set;
+	unsigned long pages;
+	int err;
+	size_t i;
+	struct landlock_prog_set tmp_prog_set = {};
+
+	if (prog->type != BPF_PROG_TYPE_LANDLOCK_HOOK)
+		return ERR_PTR(-EINVAL);
+
+	/* validate memory size allocation */
+	pages = prog->pages;
+	if (current_prog_set) {
+		size_t i;
+
+		for (i = 0; i < ARRAY_SIZE(current_prog_set->programs); i++) {
+			struct landlock_prog_list *walker_p;
+
+			for (walker_p = current_prog_set->programs[i];
+					walker_p; walker_p = walker_p->prev)
+				pages += walker_p->prog->pages;
+		}
+		/* count a struct landlock_prog_set if we need to allocate one */
+		if (refcount_read(&current_prog_set->usage) != 1)
+			pages += round_up(sizeof(*current_prog_set), PAGE_SIZE)
+				/ PAGE_SIZE;
+	}
+	if (pages > LANDLOCK_PROGRAMS_MAX_PAGES)
+		return ERR_PTR(-E2BIG);
+
+	/* ensure early that we can allocate enough memory for the new
+	 * prog_lists */
+	err = store_landlock_prog(&tmp_prog_set, current_prog_set, prog);
+	if (err)
+		return ERR_PTR(err);
+
+	/*
+	 * Each task_struct points to an array of prog list pointers.  These
+	 * tables are duplicated when additions are made (which means each
+	 * table needs to be refcounted for the processes using it). When a new
+	 * table is created, all the refcounters on the prog_list are bumped (to
+	 * track each table that references the prog). When a new prog is
+	 * added, it's just prepended to the list for the new table to point
+	 * at.
+	 *
+	 * Manage all the possible errors before this step to not uselessly
+	 * duplicate current_prog_set and avoid a rollback.
+	 */
+	if (!new_prog_set) {
+		/*
+		 * If there is no Landlock program set used by the current task,
+		 * then create a new one.
+		 */
+		new_prog_set = new_landlock_prog_set();
+		if (IS_ERR(new_prog_set))
+			goto put_tmp_lists;
+	} else if (refcount_read(&current_prog_set->usage) > 1) {
+		/*
+		 * If the current task is not the sole user of its Landlock
+		 * program set, then duplicate them.
+		 */
+		new_prog_set = new_landlock_prog_set();
+		if (IS_ERR(new_prog_set))
+			goto put_tmp_lists;
+		for (i = 0; i < ARRAY_SIZE(new_prog_set->programs); i++) {
+			new_prog_set->programs[i] =
+				READ_ONCE(current_prog_set->programs[i]);
+			if (new_prog_set->programs[i])
+				refcount_inc(&new_prog_set->programs[i]->usage);
+		}
+
+		/*
+		 * Landlock program set from the current task will not be freed
+		 * here because the usage is strictly greater than 1. It is
+		 * only prevented to be freed by another task thanks to the
+		 * caller of landlock_prepend_prog() which should be locked if
+		 * needed.
+		 */
+		landlock_put_prog_set(current_prog_set);
+	}
+
+	/* prepend tmp_prog_set to new_prog_set */
+	for (i = 0; i < ARRAY_SIZE(tmp_prog_set.programs); i++) {
+		/* get the last new list */
+		struct landlock_prog_list *last_list =
+			tmp_prog_set.programs[i];
+
+		if (last_list) {
+			while (last_list->prev)
+				last_list = last_list->prev;
+			/* no need to increment usage (pointer replacement) */
+			last_list->prev = new_prog_set->programs[i];
+			new_prog_set->programs[i] = tmp_prog_set.programs[i];
+		}
+	}
+	return new_prog_set;
+
+put_tmp_lists:
+	for (i = 0; i < ARRAY_SIZE(tmp_prog_set.programs); i++)
+		put_landlock_prog_list(tmp_prog_set.programs[i]);
+	return new_prog_set;
+}
diff --git a/security/landlock/enforce.h b/security/landlock/enforce.h
new file mode 100644
index 000000000000..39b800d9999f
--- /dev/null
+++ b/security/landlock/enforce.h
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Landlock LSM - enforcing helpers headers
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#ifndef _SECURITY_LANDLOCK_ENFORCE_H
+#define _SECURITY_LANDLOCK_ENFORCE_H
+
+struct landlock_prog_set *landlock_prepend_prog(
+		struct landlock_prog_set *current_prog_set,
+		struct bpf_prog *prog);
+void landlock_put_prog_set(struct landlock_prog_set *prog_set);
+void landlock_get_prog_set(struct landlock_prog_set *prog_set);
+
+#endif /* _SECURITY_LANDLOCK_ENFORCE_H */
diff --git a/security/landlock/enforce_seccomp.c b/security/landlock/enforce_seccomp.c
new file mode 100644
index 000000000000..c38c81e6b01a
--- /dev/null
+++ b/security/landlock/enforce_seccomp.c
@@ -0,0 +1,92 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Landlock LSM - enforcing with seccomp
+ *
+ * Copyright © 2016-2018 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#ifdef CONFIG_SECCOMP_FILTER
+
+#include <linux/bpf.h> /* bpf_prog_put() */
+#include <linux/capability.h>
+#include <linux/err.h> /* PTR_ERR() */
+#include <linux/errno.h>
+#include <linux/filter.h> /* struct bpf_prog */
+#include <linux/landlock.h>
+#include <linux/refcount.h>
+#include <linux/sched.h> /* current */
+#include <linux/uaccess.h> /* get_user() */
+
+#include "enforce.h"
+
+/* headers in include/linux/landlock.h */
+
+/**
+ * landlock_seccomp_prepend_prog - attach a Landlock program to the current
+ *                                 process
+ *
+ * current->seccomp.landlock_state->prog_set is lazily allocated. When a
+ * process fork, only a pointer is copied.  When a new program is added by a
+ * process, if there is other references to this process' prog_set, then a new
+ * allocation is made to contain an array pointing to Landlock program lists.
+ * This design enable low-performance impact and is memory efficient while
+ * keeping the property of prepend-only programs.
+ *
+ * For now, installing a Landlock prog requires that the requesting task has
+ * the global CAP_SYS_ADMIN. We cannot force the use of no_new_privs to not
+ * exclude containers where a process may legitimately acquire more privileges
+ * thanks to an SUID binary.
+ *
+ * @flags: not used for now, but could be used for TSYNC
+ * @user_bpf_fd: file descriptor pointing to a loaded Landlock prog
+ */
+int landlock_seccomp_prepend_prog(unsigned int flags,
+		const int __user *user_bpf_fd)
+{
+	struct landlock_prog_set *new_prog_set;
+	struct bpf_prog *prog;
+	int bpf_fd, err;
+
+	/* planned to be replaced with a no_new_privs check to allow
+	 * unprivileged tasks */
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+	/* enable to check if Landlock is supported with early EFAULT */
+	if (!user_bpf_fd)
+		return -EFAULT;
+	if (flags)
+		return -EINVAL;
+	err = get_user(bpf_fd, user_bpf_fd);
+	if (err)
+		return err;
+
+	prog = bpf_prog_get(bpf_fd);
+	if (IS_ERR(prog))
+		return PTR_ERR(prog);
+
+	/*
+	 * We don't need to lock anything for the current process hierarchy,
+	 * everything is guarded by the atomic counters.
+	 */
+	new_prog_set = landlock_prepend_prog(
+			current->seccomp.landlock_prog_set, prog);
+	bpf_prog_put(prog);
+	/* @prog is managed/freed by landlock_prepend_prog() */
+	if (IS_ERR(new_prog_set))
+		return PTR_ERR(new_prog_set);
+	current->seccomp.landlock_prog_set = new_prog_set;
+	return 0;
+}
+
+void put_seccomp_landlock(struct task_struct *tsk)
+{
+	landlock_put_prog_set(tsk->seccomp.landlock_prog_set);
+}
+
+void get_seccomp_landlock(struct task_struct *tsk)
+{
+	landlock_get_prog_set(tsk->seccomp.landlock_prog_set);
+}
+
+#endif /* CONFIG_SECCOMP_FILTER */
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 05/10] landlock: Handle filesystem access control
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
                   ` (3 preceding siblings ...)
  2019-07-21 21:31 ` [PATCH bpf-next v10 04/10] seccomp,landlock: Enforce Landlock programs per process hierarchy Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode Mickaël Salaün
                   ` (4 subsequent siblings)
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

This add two Landlock hooks: FS_WALK and FS_PICK.

The FS_WALK hook is used to walk through a file path. A program tied to
this hook will be evaluated for each directory traversal except the last
one if it is the leaf of the path.  It is important to differentiate
this hook from FS_PICK to enable more powerful path evaluation in the
future (cf. Landlock patch v8).

The FS_PICK hook is used to validate a set of actions requested on a
file. This actions are defined with triggers (e.g. read, write, open,
append...).

The Landlock LSM hook registration is done after other LSM to only run
actions from user-space, via eBPF programs, if the access was granted by
major (privileged) LSMs.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: James Morris <jmorris@namei.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
---

Changes since v9:
* replace subtype with expected_attach_type and expected_attach_triggers

Changes since v8:
* add a new LSM_ORDER_LAST, cf. commit e2bc445b66ca ("LSM: Introduce
  enum lsm_order")
* add WARN_ON() for pointer dereferencement
* remove the FS_GET subtype which rely on program chaining
* remove the subtype option which was only used for chaining (with the
  "previous" field)
* remove inode_lookup which depends on the (removed) nameidata security
  blob
* remove eBPF helpers to get and set Landlock inode tags
* do not use task LSM credentials (for now)

Changes since v7:
* major rewrite with clean Landlock hooks able to deal with file paths

Changes since v6:
* add 3 more sub-events: IOCTL, LOCK, FCNTL
  https://lkml.kernel.org/r/2fbc99a6-f190-f335-bd14-04bdeed35571@digikod.net
* use the new security_add_hooks()
* explain the -Werror=unused-function
* constify pointers
* cleanup headers

Changes since v5:
* split hooks.[ch] into hooks.[ch] and hooks_fs.[ch]
* add more documentation
* cosmetic fixes
* rebase (SCALAR_VALUE)

Changes since v4:
* add LSM hook abstraction called Landlock event
  * use the compiler type checking to verify hooks use by an event
  * handle all filesystem related LSM hooks (e.g. file_permission,
    mmap_file, sb_mount...)
* register BPF programs for Landlock just after LSM hooks registration
* move hooks registration after other LSMs
* add failsafes to check if a hook is not used by the kernel
* allow partial raw value access form the context (needed for programs
  generated by LLVM)

Changes since v3:
* split commit
* add hooks dealing with struct inode and struct path pointers:
  inode_permission and inode_getattr
* add abstraction over eBPF helper arguments thanks to wrapping structs
---
 include/linux/lsm_hooks.h    |   1 +
 security/landlock/Makefile   |   3 +-
 security/landlock/common.h   |   9 +
 security/landlock/hooks.c    |  94 ++++++
 security/landlock/hooks.h    |  31 ++
 security/landlock/hooks_fs.c | 554 +++++++++++++++++++++++++++++++++++
 security/landlock/hooks_fs.h |  31 ++
 security/landlock/init.c     |  33 +++
 security/security.c          |  15 +
 9 files changed, 770 insertions(+), 1 deletion(-)
 create mode 100644 security/landlock/hooks.c
 create mode 100644 security/landlock/hooks.h
 create mode 100644 security/landlock/hooks_fs.c
 create mode 100644 security/landlock/hooks_fs.h

diff --git a/include/linux/lsm_hooks.h b/include/linux/lsm_hooks.h
index df1318d85f7d..c06ad8a1d424 100644
--- a/include/linux/lsm_hooks.h
+++ b/include/linux/lsm_hooks.h
@@ -2092,6 +2092,7 @@ extern void security_add_hooks(struct security_hook_list *hooks, int count,
 enum lsm_order {
 	LSM_ORDER_FIRST = -1,	/* This is only for capabilities. */
 	LSM_ORDER_MUTABLE = 0,
+	LSM_ORDER_LAST = 1,	/* potentially-unprivileged LSM */
 };
 
 struct lsm_info {
diff --git a/security/landlock/Makefile b/security/landlock/Makefile
index 2a1a7082a365..270ece5d93de 100644
--- a/security/landlock/Makefile
+++ b/security/landlock/Makefile
@@ -1,4 +1,5 @@
 obj-$(CONFIG_SECURITY_LANDLOCK) := landlock.o
 
 landlock-y := init.o \
-	enforce.o enforce_seccomp.o
+	enforce.o enforce_seccomp.o \
+	hooks.o hooks_fs.o
diff --git a/security/landlock/common.h b/security/landlock/common.h
index 2cf36dbf4560..b2ee018eb6fc 100644
--- a/security/landlock/common.h
+++ b/security/landlock/common.h
@@ -79,4 +79,13 @@ static inline enum landlock_hook_type get_hook_type(const struct bpf_prog *prog)
 	}
 }
 
+__maybe_unused
+static bool current_has_prog_type(enum landlock_hook_type hook_type)
+{
+	struct landlock_prog_set *prog_set;
+
+	prog_set = current->seccomp.landlock_prog_set;
+	return (prog_set && prog_set->programs[get_hook_index(hook_type)]);
+}
+
 #endif /* _SECURITY_LANDLOCK_COMMON_H */
diff --git a/security/landlock/hooks.c b/security/landlock/hooks.c
new file mode 100644
index 000000000000..97c54957f17b
--- /dev/null
+++ b/security/landlock/hooks.c
@@ -0,0 +1,94 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Landlock LSM - hook helpers
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#include <asm/current.h>
+#include <linux/bpf.h> /* enum bpf_prog_aux */
+#include <linux/errno.h>
+#include <linux/filter.h> /* BPF_PROG_RUN() */
+#include <linux/rculist.h> /* list_add_tail_rcu */
+#include <uapi/linux/landlock.h> /* struct landlock_context */
+
+#include "common.h" /* struct landlock_rule, get_hook_index() */
+#include "hooks.h" /* landlock_hook_ctx */
+
+#include "hooks_fs.h"
+
+/* return a Landlock program context (e.g. hook_ctx->fs_walk.prog_ctx) */
+static const void *get_ctx(enum landlock_hook_type hook_type,
+		struct landlock_hook_ctx *hook_ctx)
+{
+	switch (hook_type) {
+	case LANDLOCK_HOOK_FS_WALK:
+		return landlock_get_ctx_fs_walk(hook_ctx->fs_walk);
+	case LANDLOCK_HOOK_FS_PICK:
+		return landlock_get_ctx_fs_pick(hook_ctx->fs_pick);
+	}
+	WARN_ON(1);
+	return NULL;
+}
+
+/**
+ * landlock_access_deny - run Landlock programs tied to a hook
+ *
+ * @hook_idx: hook index in the programs array
+ * @ctx: non-NULL valid eBPF context
+ * @prog_set: Landlock program set pointer
+ * @triggers: a bitmask to check if a program should be run
+ *
+ * Return true if at least one program return deny.
+ */
+static bool landlock_access_deny(enum landlock_hook_type hook_type,
+		struct landlock_hook_ctx *hook_ctx,
+		struct landlock_prog_set *prog_set, u64 triggers)
+{
+	struct landlock_prog_list *prog_list, *prev_list = NULL;
+	u32 hook_idx = get_hook_index(hook_type);
+
+	if (!prog_set)
+		return false;
+
+	for (prog_list = prog_set->programs[hook_idx];
+			prog_list; prog_list = prog_list->prev) {
+		u32 ret;
+		const void *prog_ctx;
+
+		/* check if @prog expect at least one of this triggers */
+		if (triggers && !(triggers & prog_list->prog->aux->
+					expected_attach_triggers))
+			continue;
+		prog_ctx = get_ctx(hook_type, hook_ctx);
+		if (!prog_ctx || WARN_ON(IS_ERR(prog_ctx)))
+			return true;
+		rcu_read_lock();
+		ret = BPF_PROG_RUN(prog_list->prog, prog_ctx);
+		rcu_read_unlock();
+		/* deny access if a program returns a value different than 0 */
+		if (ret)
+			return true;
+		if (prev_list && prog_list->prev && prog_list->prev->prog->
+				expected_attach_type ==
+				prev_list->prog->expected_attach_type)
+			WARN_ON(prog_list->prev != prev_list);
+		prev_list = prog_list;
+	}
+	return false;
+}
+
+int landlock_decide(enum landlock_hook_type hook_type,
+		struct landlock_hook_ctx *hook_ctx, u64 triggers)
+{
+	bool deny = false;
+
+#ifdef CONFIG_SECCOMP_FILTER
+	deny = landlock_access_deny(hook_type, hook_ctx,
+			current->seccomp.landlock_prog_set, triggers);
+#endif /* CONFIG_SECCOMP_FILTER */
+
+	/* should we use -EPERM or -EACCES? */
+	return deny ? -EACCES : 0;
+}
diff --git a/security/landlock/hooks.h b/security/landlock/hooks.h
new file mode 100644
index 000000000000..31446e6629fb
--- /dev/null
+++ b/security/landlock/hooks.h
@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Landlock LSM - hooks helpers
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#include <asm/current.h>
+#include <linux/sched.h> /* struct task_struct */
+#include <linux/seccomp.h>
+
+#include "hooks_fs.h"
+
+struct landlock_hook_ctx {
+	union {
+		struct landlock_hook_ctx_fs_walk *fs_walk;
+		struct landlock_hook_ctx_fs_pick *fs_pick;
+	};
+};
+
+static inline bool landlocked(const struct task_struct *task)
+{
+#ifdef CONFIG_SECCOMP_FILTER
+	return !!(task->seccomp.landlock_prog_set);
+#else
+	return false;
+#endif /* CONFIG_SECCOMP_FILTER */
+}
+
+int landlock_decide(enum landlock_hook_type, struct landlock_hook_ctx *, u64);
diff --git a/security/landlock/hooks_fs.c b/security/landlock/hooks_fs.c
new file mode 100644
index 000000000000..3f81b7fc2938
--- /dev/null
+++ b/security/landlock/hooks_fs.c
@@ -0,0 +1,554 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Landlock LSM - filesystem hooks
+ *
+ * Copyright © 2016-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#include <linux/bpf.h> /* enum bpf_access_type */
+#include <linux/kernel.h> /* ARRAY_SIZE */
+#include <linux/lsm_hooks.h>
+#include <linux/rcupdate.h> /* synchronize_rcu() */
+#include <linux/stat.h> /* S_ISDIR */
+#include <linux/stddef.h> /* offsetof */
+#include <linux/types.h> /* uintptr_t */
+#include <linux/workqueue.h> /* INIT_WORK() */
+
+/* permissions translation */
+#include <linux/fs.h> /* MAY_* */
+#include <linux/mman.h> /* PROT_* */
+#include <linux/namei.h>
+
+/* hook arguments */
+#include <linux/dcache.h> /* struct dentry */
+#include <linux/fs.h> /* struct inode, struct iattr */
+#include <linux/mm_types.h> /* struct vm_area_struct */
+#include <linux/mount.h> /* struct vfsmount */
+#include <linux/path.h> /* struct path */
+#include <linux/sched.h> /* struct task_struct */
+#include <linux/time.h> /* struct timespec */
+
+#include "common.h"
+#include "hooks_fs.h"
+#include "hooks.h"
+
+/* fs_pick */
+
+#include <asm/page.h> /* PAGE_SIZE */
+#include <asm/syscall.h>
+#include <linux/dcache.h> /* d_path, dentry_path_raw */
+#include <linux/err.h> /* *_ERR */
+#include <linux/gfp.h> /* __get_free_page, GFP_KERNEL */
+#include <linux/path.h> /* struct path */
+
+bool landlock_is_valid_access_fs_pick(int off, enum bpf_access_type type,
+		enum bpf_reg_type *reg_type, int *max_size)
+{
+	switch (off) {
+	default:
+		return false;
+	}
+}
+
+bool landlock_is_valid_access_fs_walk(int off, enum bpf_access_type type,
+		enum bpf_reg_type *reg_type, int *max_size)
+{
+	switch (off) {
+	default:
+		return false;
+	}
+}
+
+/* fs_walk */
+
+struct landlock_hook_ctx_fs_walk {
+	struct landlock_ctx_fs_walk prog_ctx;
+};
+
+const struct landlock_ctx_fs_walk *landlock_get_ctx_fs_walk(
+		const struct landlock_hook_ctx_fs_walk *hook_ctx)
+{
+	if (WARN_ON(!hook_ctx))
+		return NULL;
+
+	return &hook_ctx->prog_ctx;
+}
+
+static int decide_fs_walk(int may_mask, struct inode *inode)
+{
+	struct landlock_hook_ctx_fs_walk fs_walk = {};
+	struct landlock_hook_ctx hook_ctx = {
+		.fs_walk = &fs_walk,
+	};
+	const enum landlock_hook_type hook_type = LANDLOCK_HOOK_FS_WALK;
+
+	if (!current_has_prog_type(hook_type))
+		/* no fs_walk */
+		return 0;
+	if (WARN_ON(!inode))
+		return -EFAULT;
+
+	/* init common data: inode */
+	fs_walk.prog_ctx.inode = (uintptr_t)inode;
+	return landlock_decide(hook_type, &hook_ctx, 0);
+}
+
+/* fs_pick */
+
+struct landlock_hook_ctx_fs_pick {
+	__u64 triggers;
+	struct landlock_ctx_fs_pick prog_ctx;
+};
+
+const struct landlock_ctx_fs_pick *landlock_get_ctx_fs_pick(
+		const struct landlock_hook_ctx_fs_pick *hook_ctx)
+{
+	if (WARN_ON(!hook_ctx))
+		return NULL;
+
+	return &hook_ctx->prog_ctx;
+}
+
+static int decide_fs_pick(__u64 triggers, struct inode *inode)
+{
+	struct landlock_hook_ctx_fs_pick fs_pick = {};
+	struct landlock_hook_ctx hook_ctx = {
+		.fs_pick = &fs_pick,
+	};
+	const enum landlock_hook_type hook_type = LANDLOCK_HOOK_FS_PICK;
+
+	if (WARN_ON(!triggers))
+		return 0;
+	if (!current_has_prog_type(hook_type))
+		/* no fs_pick */
+		return 0;
+	if (WARN_ON(!inode))
+		return -EFAULT;
+
+	fs_pick.triggers = triggers,
+	/* init common data: inode */
+	fs_pick.prog_ctx.inode = (uintptr_t)inode;
+	return landlock_decide(hook_type, &hook_ctx, fs_pick.triggers);
+}
+
+/* helpers */
+
+static u64 fs_may_to_triggers(int may_mask, umode_t mode)
+{
+	u64 ret = 0;
+
+	if (may_mask & MAY_EXEC)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_EXECUTE;
+	if (may_mask & MAY_READ) {
+		if (S_ISDIR(mode))
+			ret |= LANDLOCK_TRIGGER_FS_PICK_READDIR;
+		else
+			ret |= LANDLOCK_TRIGGER_FS_PICK_READ;
+	}
+	if (may_mask & MAY_WRITE)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_WRITE;
+	if (may_mask & MAY_APPEND)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_APPEND;
+	if (may_mask & MAY_OPEN)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_OPEN;
+	if (may_mask & MAY_CHROOT)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_CHROOT;
+	else if (may_mask & MAY_CHDIR)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_CHDIR;
+	/* XXX: ignore MAY_ACCESS */
+	WARN_ON(!ret);
+	return ret;
+}
+
+static inline u64 mem_prot_to_triggers(unsigned long prot, bool private)
+{
+	u64 ret = LANDLOCK_TRIGGER_FS_PICK_MAP;
+
+	/* private mapping do not write to files */
+	if (!private && (prot & PROT_WRITE))
+		ret |= LANDLOCK_TRIGGER_FS_PICK_WRITE;
+	if (prot & PROT_READ)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_READ;
+	if (prot & PROT_EXEC)
+		ret |= LANDLOCK_TRIGGER_FS_PICK_EXECUTE;
+	WARN_ON(!ret);
+	return ret;
+}
+
+/* binder hooks */
+
+static int hook_binder_transfer_file(struct task_struct *from,
+		struct task_struct *to, struct file *file)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!file))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_TRANSFER,
+			file_inode(file));
+}
+
+/* sb hooks */
+
+static int hook_sb_statfs(struct dentry *dentry)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_GETATTR,
+			dentry->d_inode);
+}
+
+/* TODO: handle mount source and remount */
+static int hook_sb_mount(const char *dev_name, const struct path *path,
+		const char *type, unsigned long flags, void *data)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!path))
+		return 0;
+	if (WARN_ON(!path->dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_MOUNTON,
+			path->dentry->d_inode);
+}
+
+/*
+ * The @old_path is similar to a destination mount point.
+ */
+static int hook_sb_pivotroot(const struct path *old_path,
+		const struct path *new_path)
+{
+	int err;
+
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!old_path))
+		return 0;
+	if (WARN_ON(!old_path->dentry))
+		return 0;
+	err = decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_MOUNTON,
+			old_path->dentry->d_inode);
+	if (err)
+		return err;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_CHROOT,
+			new_path->dentry->d_inode);
+}
+
+/* inode hooks */
+
+/* a directory inode contains only one dentry */
+static int hook_inode_create(struct inode *dir, struct dentry *dentry,
+		umode_t mode)
+{
+	if (!landlocked(current))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_CREATE, dir);
+}
+
+static int hook_inode_link(struct dentry *old_dentry, struct inode *dir,
+		struct dentry *new_dentry)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!old_dentry)) {
+		int ret = decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_LINK,
+				old_dentry->d_inode);
+		if (ret)
+			return ret;
+	}
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_LINKTO, dir);
+}
+
+static int hook_inode_unlink(struct inode *dir, struct dentry *dentry)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_UNLINK,
+			dentry->d_inode);
+}
+
+static int hook_inode_symlink(struct inode *dir, struct dentry *dentry,
+		const char *old_name)
+{
+	if (!landlocked(current))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_CREATE, dir);
+}
+
+static int hook_inode_mkdir(struct inode *dir, struct dentry *dentry,
+		umode_t mode)
+{
+	if (!landlocked(current))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_CREATE, dir);
+}
+
+static int hook_inode_rmdir(struct inode *dir, struct dentry *dentry)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_RMDIR, dentry->d_inode);
+}
+
+static int hook_inode_mknod(struct inode *dir, struct dentry *dentry,
+		umode_t mode, dev_t dev)
+{
+	if (!landlocked(current))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_CREATE, dir);
+}
+
+static int hook_inode_rename(struct inode *old_dir, struct dentry *old_dentry,
+		struct inode *new_dir, struct dentry *new_dentry)
+{
+	if (!landlocked(current))
+		return 0;
+	if (!WARN_ON(!old_dentry)) {
+		int ret = decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_RENAME,
+				old_dentry->d_inode);
+		if (ret)
+			return ret;
+	}
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_RENAMETO, new_dir);
+}
+
+static int hook_inode_readlink(struct dentry *dentry)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_READ, dentry->d_inode);
+}
+
+/*
+ * ignore the inode_follow_link hook (could set is_symlink in the fs_walk
+ * context)
+ */
+
+static int hook_inode_permission(struct inode *inode, int mask)
+{
+	u64 triggers;
+
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!inode))
+		return 0;
+
+	triggers = fs_may_to_triggers(mask, inode->i_mode);
+	/*
+	 * decide_fs_walk() is exclusive with decide_fs_pick(): in a path walk,
+	 * ignore execute-only access on directory for any fs_pick program
+	 */
+	if (triggers == LANDLOCK_TRIGGER_FS_PICK_EXECUTE &&
+			S_ISDIR(inode->i_mode))
+		return decide_fs_walk(mask, inode);
+
+	return decide_fs_pick(triggers, inode);
+}
+
+static int hook_inode_setattr(struct dentry *dentry, struct iattr *attr)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_SETATTR,
+			dentry->d_inode);
+}
+
+static int hook_inode_getattr(const struct path *path)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!path))
+		return 0;
+	if (WARN_ON(!path->dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_GETATTR,
+			path->dentry->d_inode);
+}
+
+static int hook_inode_setxattr(struct dentry *dentry, const char *name,
+		const void *value, size_t size, int flags)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_SETATTR,
+			dentry->d_inode);
+}
+
+static int hook_inode_getxattr(struct dentry *dentry, const char *name)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_GETATTR,
+			dentry->d_inode);
+}
+
+static int hook_inode_listxattr(struct dentry *dentry)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_GETATTR,
+			dentry->d_inode);
+}
+
+static int hook_inode_removexattr(struct dentry *dentry, const char *name)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!dentry))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_SETATTR,
+			dentry->d_inode);
+}
+
+static int hook_inode_getsecurity(struct inode *inode, const char *name,
+		void **buffer, bool alloc)
+{
+	if (!landlocked(current))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_GETATTR, inode);
+}
+
+static int hook_inode_setsecurity(struct inode *inode, const char *name,
+		const void *value, size_t size, int flag)
+{
+	if (!landlocked(current))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_SETATTR, inode);
+}
+
+static int hook_inode_listsecurity(struct inode *inode, char *buffer,
+		size_t buffer_size)
+{
+	if (!landlocked(current))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_GETATTR, inode);
+}
+
+/* file hooks */
+
+static int hook_file_ioctl(struct file *file, unsigned int cmd,
+		unsigned long arg)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!file))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_IOCTL,
+			file_inode(file));
+}
+
+static int hook_file_lock(struct file *file, unsigned int cmd)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!file))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_LOCK, file_inode(file));
+}
+
+static int hook_file_fcntl(struct file *file, unsigned int cmd,
+		unsigned long arg)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!file))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_FCNTL,
+			file_inode(file));
+}
+
+static int hook_mmap_file(struct file *file, unsigned long reqprot,
+		unsigned long prot, unsigned long flags)
+{
+	if (!landlocked(current))
+		return 0;
+	/* file can be null for anonymous mmap */
+	if (!file)
+		return 0;
+	return decide_fs_pick(mem_prot_to_triggers(prot, flags & MAP_PRIVATE),
+			file_inode(file));
+}
+
+static int hook_file_mprotect(struct vm_area_struct *vma,
+		unsigned long reqprot, unsigned long prot)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!vma))
+		return 0;
+	if (!vma->vm_file)
+		return 0;
+	return decide_fs_pick(mem_prot_to_triggers(prot,
+				!(vma->vm_flags & VM_SHARED)),
+			file_inode(vma->vm_file));
+}
+
+static int hook_file_receive(struct file *file)
+{
+	if (!landlocked(current))
+		return 0;
+	if (WARN_ON(!file))
+		return 0;
+	return decide_fs_pick(LANDLOCK_TRIGGER_FS_PICK_RECEIVE,
+			file_inode(file));
+}
+
+static struct security_hook_list landlock_hooks[] = {
+	LSM_HOOK_INIT(binder_transfer_file, hook_binder_transfer_file),
+
+	LSM_HOOK_INIT(sb_statfs, hook_sb_statfs),
+	LSM_HOOK_INIT(sb_mount, hook_sb_mount),
+	LSM_HOOK_INIT(sb_pivotroot, hook_sb_pivotroot),
+
+	LSM_HOOK_INIT(inode_create, hook_inode_create),
+	LSM_HOOK_INIT(inode_link, hook_inode_link),
+	LSM_HOOK_INIT(inode_unlink, hook_inode_unlink),
+	LSM_HOOK_INIT(inode_symlink, hook_inode_symlink),
+	LSM_HOOK_INIT(inode_mkdir, hook_inode_mkdir),
+	LSM_HOOK_INIT(inode_rmdir, hook_inode_rmdir),
+	LSM_HOOK_INIT(inode_mknod, hook_inode_mknod),
+	LSM_HOOK_INIT(inode_rename, hook_inode_rename),
+	LSM_HOOK_INIT(inode_readlink, hook_inode_readlink),
+	LSM_HOOK_INIT(inode_permission, hook_inode_permission),
+	LSM_HOOK_INIT(inode_setattr, hook_inode_setattr),
+	LSM_HOOK_INIT(inode_getattr, hook_inode_getattr),
+	LSM_HOOK_INIT(inode_setxattr, hook_inode_setxattr),
+	LSM_HOOK_INIT(inode_getxattr, hook_inode_getxattr),
+	LSM_HOOK_INIT(inode_listxattr, hook_inode_listxattr),
+	LSM_HOOK_INIT(inode_removexattr, hook_inode_removexattr),
+	LSM_HOOK_INIT(inode_getsecurity, hook_inode_getsecurity),
+	LSM_HOOK_INIT(inode_setsecurity, hook_inode_setsecurity),
+	LSM_HOOK_INIT(inode_listsecurity, hook_inode_listsecurity),
+
+	/* do not handle file_permission for now */
+	LSM_HOOK_INIT(file_ioctl, hook_file_ioctl),
+	LSM_HOOK_INIT(file_lock, hook_file_lock),
+	LSM_HOOK_INIT(file_fcntl, hook_file_fcntl),
+	LSM_HOOK_INIT(mmap_file, hook_mmap_file),
+	LSM_HOOK_INIT(file_mprotect, hook_file_mprotect),
+	LSM_HOOK_INIT(file_receive, hook_file_receive),
+	/* file_open is not handled, use inode_permission instead */
+};
+
+__init void landlock_add_hooks_fs(void)
+{
+	security_add_hooks(landlock_hooks, ARRAY_SIZE(landlock_hooks),
+			LANDLOCK_NAME);
+}
diff --git a/security/landlock/hooks_fs.h b/security/landlock/hooks_fs.h
new file mode 100644
index 000000000000..eeae4dcd842f
--- /dev/null
+++ b/security/landlock/hooks_fs.h
@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Landlock LSM - filesystem hooks
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2018-2019 ANSSI
+ */
+
+#include <linux/bpf.h> /* enum bpf_access_type */
+
+__init void landlock_add_hooks_fs(void);
+
+/* fs_pick */
+
+struct landlock_hook_ctx_fs_pick;
+
+bool landlock_is_valid_access_fs_pick(int off, enum bpf_access_type type,
+		enum bpf_reg_type *reg_type, int *max_size);
+
+const struct landlock_ctx_fs_pick *landlock_get_ctx_fs_pick(
+		const struct landlock_hook_ctx_fs_pick *hook_ctx);
+
+/* fs_walk */
+
+struct landlock_hook_ctx_fs_walk;
+
+bool landlock_is_valid_access_fs_walk(int off, enum bpf_access_type type,
+		enum bpf_reg_type *reg_type, int *max_size);
+
+const struct landlock_ctx_fs_walk *landlock_get_ctx_fs_walk(
+		const struct landlock_hook_ctx_fs_walk *hook_ctx);
diff --git a/security/landlock/init.c b/security/landlock/init.c
index 8dfd5fea3c1f..391e88bd4d3a 100644
--- a/security/landlock/init.c
+++ b/security/landlock/init.c
@@ -9,8 +9,10 @@
 #include <linux/bpf.h> /* enum bpf_access_type */
 #include <linux/capability.h> /* capable */
 #include <linux/filter.h> /* struct bpf_prog */
+#include <linux/lsm_hooks.h>
 
 #include "common.h" /* LANDLOCK_* */
+#include "hooks_fs.h"
 
 static bool bpf_landlock_is_valid_access(int off, int size,
 		enum bpf_access_type type, const struct bpf_prog *prog,
@@ -27,6 +29,20 @@ static bool bpf_landlock_is_valid_access(int off, int size,
 	if (size <= 0 || size > sizeof(__u64))
 		return false;
 
+	/* set register type and max size */
+	switch (get_hook_type(prog)) {
+	case LANDLOCK_HOOK_FS_PICK:
+		if (!landlock_is_valid_access_fs_pick(off, type, &reg_type,
+					&max_size))
+			return false;
+		break;
+	case LANDLOCK_HOOK_FS_WALK:
+		if (!landlock_is_valid_access_fs_walk(off, type, &reg_type,
+					&max_size))
+			return false;
+		break;
+	}
+
 	/* check memory range access */
 	switch (reg_type) {
 	case NOT_INIT:
@@ -98,3 +114,20 @@ const struct bpf_verifier_ops landlock_verifier_ops = {
 };
 
 const struct bpf_prog_ops landlock_prog_ops = {};
+
+static int __init landlock_init(void)
+{
+	pr_info(LANDLOCK_NAME ": Initializing (sandbox with seccomp)\n");
+	landlock_add_hooks_fs();
+	return 0;
+}
+
+struct lsm_blob_sizes landlock_blob_sizes __lsm_ro_after_init = {
+};
+
+DEFINE_LSM(LANDLOCK_NAME) = {
+	.name = LANDLOCK_NAME,
+	.order = LSM_ORDER_LAST,
+	.blobs = &landlock_blob_sizes,
+	.init = landlock_init,
+};
diff --git a/security/security.c b/security/security.c
index 250ee2d76406..e694e5fe7021 100644
--- a/security/security.c
+++ b/security/security.c
@@ -263,6 +263,21 @@ static void __init ordered_lsm_parse(const char *order, const char *origin)
 		}
 	}
 
+	/*
+	 * In case of an unprivileged access-control, we don't want to give the
+	 * ability to any process to do some checks (e.g. through an eBPF
+	 * program) on kernel objects (e.g. files) if a privileged security
+	 * policy forbid their access.  We must then load
+	 * potentially-unprivileged security modules after all other LSMs.
+	 *
+	 * LSM_ORDER_LAST is always last and does not appear in the modifiable
+	 * ordered list of enabled LSMs.
+	 */
+	for (lsm = __start_lsm_info; lsm < __end_lsm_info; lsm++) {
+		if (lsm->order == LSM_ORDER_LAST)
+			append_ordered_lsm(lsm, "last");
+	}
+
 	/* Disable all LSMs not in the ordered list. */
 	for (lsm = __start_lsm_info; lsm < __end_lsm_info; lsm++) {
 		if (exists_ordered_lsm(lsm))
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
                   ` (4 preceding siblings ...)
  2019-07-21 21:31 ` [PATCH bpf-next v10 05/10] landlock: Handle filesystem access control Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-27  1:40   ` Alexei Starovoitov
  2019-07-21 21:31 ` [PATCH bpf-next v10 07/10] landlock: Add ptrace restrictions Mickaël Salaün
                   ` (3 subsequent siblings)
  9 siblings, 1 reply; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

FIXME: 64-bits in the doc

This new map store arbitrary values referenced by inode keys.  The map
can be updated from user space with file descriptor pointing to inodes
tied to a file system.  From an eBPF (Landlock) program point of view,
such a map is read-only and can only be used to retrieved a value tied
to a given inode.  This is useful to recognize an inode tagged by user
space, without access right to this inode (i.e. no need to have a write
access to this inode).

Add dedicated BPF functions to handle this type of map:
* bpf_inode_htab_map_update_elem()
* bpf_inode_htab_map_lookup_elem()
* bpf_inode_htab_map_delete_elem()

This new map require a dedicated helper inode_map_lookup_elem() because
of the key which is a pointer to an opaque data (only provided by the
kernel).  This act like a (physical or cryptographic) key, which is why
it is also not allowed to get the next key.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: James Morris <jmorris@namei.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
Cc: Jann Horn <jann@thejh.net>
---

Changes since v9:
* use a hash map for the inode map: integrate inodemap.c into hashtab.c
  * add map_put_key() to struct bpf_map_ops to enable to put an inode
    reference used as key
  * allow arbitrary value size instead of 64-bits
* handle inode and map lifetime with LSM hooks
* check access for inode lookup via syscall: similar to adding xattr,
  except it does not touch the file system (which is handy for read-only
  ones)
* force read-only inode map for Landlock programs
* rename inode_map_lookup() into inode_map_lookup_elem()
* fix inode and mnt checks (suggested by Al Viro)

Changes since v8:
* remove prog chaining and object tagging to ease review
* use bpf_map_init_from_attr()

Changes since v7:
* new design with a dedicated map and a BPF function to tie a value to
  an inode
* add the ability to set or get a tag on an inode from a Landlock
  program

Changes since v6:
* remove WARN_ON() for missing dentry->d_inode
* refactor bpf_landlock_func_proto() (suggested by Kees Cook)

Changes since v5:
* cosmetic fixes and rebase

Changes since v4:
* use a file abstraction (handle) to wrap inode, dentry, path and file
  structs
* remove bpf_landlock_cmp_fs_beneath()
* rename the BPF helper and move it to kernel/bpf/
* tighten helpers accessible by a Landlock rule

Changes since v3:
* remove bpf_landlock_cmp_fs_prop() (suggested by Alexei Starovoitov)
* add hooks dealing with struct inode and struct path pointers:
  inode_permission and inode_getattr
* add abstraction over eBPF helper arguments thanks to wrapping structs
* add bpf_landlock_get_fs_mode() helper to check file type and mode
* merge WARN_ON() (suggested by Kees Cook)
* fix and update bpf_helpers.h
* use BPF_CALL_* for eBPF helpers (suggested by Alexei Starovoitov)
* make handle arraymap safe (RCU) and remove buggy synchronize_rcu()
* factor out the arraymay walk
* use size_t to index array (suggested by Jann Horn)

Changes since v2:
* add MNT_INTERNAL check to only add file handle from user-visible FS
  (e.g. no anonymous inode)
* replace struct file* with struct path* in map_landlock_handle
* add BPF protos
* fix bpf_landlock_cmp_fs_prop_with_struct_file()
---
 include/linux/bpf.h            |  16 +++
 include/linux/bpf_types.h      |   3 +
 include/linux/landlock.h       |   4 +
 include/uapi/linux/bpf.h       |  12 +-
 kernel/bpf/core.c              |   2 +
 kernel/bpf/hashtab.c           | 253 +++++++++++++++++++++++++++++++++
 kernel/bpf/syscall.c           |  27 +++-
 kernel/bpf/verifier.c          |  14 ++
 security/landlock/common.h     |  14 ++
 security/landlock/hooks_fs.c   |  85 +++++++++++
 security/landlock/init.c       |  13 ++
 tools/include/uapi/linux/bpf.h |  12 +-
 tools/lib/bpf/libbpf_probes.c  |   1 +
 13 files changed, 453 insertions(+), 3 deletions(-)

diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index 6d9c7a08713e..c507438e56b5 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -47,6 +47,7 @@ struct bpf_map_ops {
 	void *(*map_fd_get_ptr)(struct bpf_map *map, struct file *map_file,
 				int fd);
 	void (*map_fd_put_ptr)(void *ptr);
+	void (*map_put_key)(void *key);
 	u32 (*map_gen_lookup)(struct bpf_map *map, struct bpf_insn *insn_buf);
 	u32 (*map_fd_sys_lookup_elem)(void *ptr);
 	void (*map_seq_show_elem)(struct bpf_map *map, void *key,
@@ -208,6 +209,8 @@ enum bpf_arg_type {
 	ARG_PTR_TO_INT,		/* pointer to int */
 	ARG_PTR_TO_LONG,	/* pointer to long */
 	ARG_PTR_TO_SOCKET,	/* pointer to bpf_sock (fullsock) */
+
+	ARG_PTR_TO_INODE,	/* pointer to a struct inode */
 };
 
 /* type of values returned from helper functions */
@@ -278,6 +281,7 @@ enum bpf_reg_type {
 	PTR_TO_TCP_SOCK_OR_NULL, /* reg points to struct tcp_sock or NULL */
 	PTR_TO_TP_BUFFER,	 /* reg points to a writable raw tp's buffer */
 	PTR_TO_XDP_SOCK,	 /* reg points to struct xdp_sock */
+	PTR_TO_INODE,		 /* reg points to struct inode */
 };
 
 /* The information passed from prog-specific *_is_valid_access
@@ -479,6 +483,7 @@ struct bpf_event_entry {
 	struct rcu_head rcu;
 };
 
+
 bool bpf_prog_array_compatible(struct bpf_array *array, const struct bpf_prog *fp);
 int bpf_prog_calc_tag(struct bpf_prog *fp);
 
@@ -684,6 +689,16 @@ int bpf_fd_array_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
 int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file,
 				void *key, void *value, u64 map_flags);
 int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
+int bpf_inode_fd_htab_map_lookup_elem(struct bpf_map *map, int *key, void *value);
+int bpf_inode_fd_htab_map_delete_elem(struct bpf_map *map, int *key);
+int bpf_inode_ptr_unlocked_htab_map_delete_elem(struct bpf_map *map,
+						struct inode **key,
+						bool remove_in_inode);
+int bpf_inode_ptr_locked_htab_map_delete_elem(struct bpf_map *map,
+					      struct inode **key,
+					      bool remove_in_inode);
+int bpf_inode_fd_htab_map_update_elem(struct bpf_map *map, int *key,
+				      void *value, u64 map_flags);
 
 int bpf_get_file_flag(int flags);
 int bpf_check_uarg_tail_zero(void __user *uaddr, size_t expected_size,
@@ -1055,6 +1070,7 @@ extern const struct bpf_func_proto bpf_get_local_storage_proto;
 extern const struct bpf_func_proto bpf_strtol_proto;
 extern const struct bpf_func_proto bpf_strtoul_proto;
 extern const struct bpf_func_proto bpf_tcp_sock_proto;
+extern const struct bpf_func_proto bpf_inode_map_lookup_elem_proto;
 
 /* Shared helpers among cBPF and eBPF. */
 void bpf_user_rnd_init_once(void);
diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h
index 2ab647323f3a..ea177818d67e 100644
--- a/include/linux/bpf_types.h
+++ b/include/linux/bpf_types.h
@@ -80,3 +80,6 @@ BPF_MAP_TYPE(BPF_MAP_TYPE_REUSEPORT_SOCKARRAY, reuseport_array_ops)
 #endif
 BPF_MAP_TYPE(BPF_MAP_TYPE_QUEUE, queue_map_ops)
 BPF_MAP_TYPE(BPF_MAP_TYPE_STACK, stack_map_ops)
+#ifdef CONFIG_SECURITY_LANDLOCK
+BPF_MAP_TYPE(BPF_MAP_TYPE_INODE, htab_inode_ops)
+#endif
diff --git a/include/linux/landlock.h b/include/linux/landlock.h
index 8ac7942f50fc..731b89cdf977 100644
--- a/include/linux/landlock.h
+++ b/include/linux/landlock.h
@@ -9,6 +9,7 @@
 #ifndef _LINUX_LANDLOCK_H
 #define _LINUX_LANDLOCK_H
 
+#include <linux/bpf.h>
 #include <linux/errno.h>
 #include <linux/sched.h> /* task_struct */
 
@@ -31,4 +32,7 @@ static inline void get_seccomp_landlock(struct task_struct *tsk)
 }
 #endif /* CONFIG_SECCOMP_FILTER && CONFIG_SECURITY_LANDLOCK */
 
+int landlock_inode_add_map(struct inode *inode, struct bpf_map *map);
+void landlock_inode_remove_map(struct inode *inode, const struct bpf_map *map);
+
 #endif /* _LINUX_LANDLOCK_H */
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index d68613f737f3..2da054ca9c8b 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -134,6 +134,7 @@ enum bpf_map_type {
 	BPF_MAP_TYPE_QUEUE,
 	BPF_MAP_TYPE_STACK,
 	BPF_MAP_TYPE_SK_STORAGE,
+	BPF_MAP_TYPE_INODE,
 };
 
 /* Note that tracing related programs such as
@@ -2717,6 +2718,14 @@ union bpf_attr {
  *		**-EPERM** if no permission to send the *sig*.
  *
  *		**-EAGAIN** if bpf program can try again.
+ *
+ * void *bpf_inode_map_lookup_elem(struct bpf_map *map, const void *key)
+ *	Description
+ *		Perform a lookup in *map* for an entry associated to an inode
+ *		*key*.
+ *	Return
+ *		Map value associated to *key*, or **NULL** if no entry was
+ *		found.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
@@ -2828,7 +2837,8 @@ union bpf_attr {
 	FN(strtoul),			\
 	FN(sk_storage_get),		\
 	FN(sk_storage_delete),		\
-	FN(send_signal),
+	FN(send_signal),		\
+	FN(inode_map_lookup_elem),
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
  * function eBPF program intends to call
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index 16079550db6d..4177c818e5cd 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -2040,6 +2040,8 @@ const struct bpf_func_proto bpf_get_current_comm_proto __weak;
 const struct bpf_func_proto bpf_get_current_cgroup_id_proto __weak;
 const struct bpf_func_proto bpf_get_local_storage_proto __weak;
 
+const struct bpf_func_proto bpf_inode_map_update_proto __weak;
+
 const struct bpf_func_proto * __weak bpf_get_trace_printk_proto(void)
 {
 	return NULL;
diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
index 22066a62c8c9..4fc7755042f0 100644
--- a/kernel/bpf/hashtab.c
+++ b/kernel/bpf/hashtab.c
@@ -1,13 +1,21 @@
 // SPDX-License-Identifier: GPL-2.0-only
 /* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
  * Copyright (c) 2016 Facebook
+ * Copyright (c) 2017-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright (c) 2019 ANSSI
  */
+#include <asm/resource.h> /* RLIMIT_NOFILE */
 #include <linux/bpf.h>
 #include <linux/btf.h>
+#include <linux/err.h>
 #include <linux/jhash.h>
+#include <linux/fs.h> /* iput() */
 #include <linux/filter.h>
+#include <linux/landlock.h>
+#include <linux/mount.h> /* MNT_INTERNAL */
 #include <linux/rculist_nulls.h>
 #include <linux/random.h>
+#include <linux/sched/signal.h> /* rlimit() */
 #include <uapi/linux/btf.h>
 #include "percpu_freelist.h"
 #include "bpf_lru_list.h"
@@ -684,6 +692,8 @@ static void free_htab_elem(struct bpf_htab *htab, struct htab_elem *l)
 
 		map->ops->map_fd_put_ptr(ptr);
 	}
+	if (map->ops->map_put_key)
+		map->ops->map_put_key(l->key);
 
 	if (htab_is_prealloc(htab)) {
 		__pcpu_freelist_push(&htab->freelist, &l->fnode);
@@ -1514,3 +1524,246 @@ const struct bpf_map_ops htab_of_maps_map_ops = {
 	.map_gen_lookup = htab_of_map_gen_lookup,
 	.map_check_btf = map_check_no_btf,
 };
+
+/* inode_htab */
+
+static int inode_htab_map_alloc_check(union bpf_attr *attr)
+{
+	/* only allow root to create this type of map (for now), should be
+	 * removed when Landlock will be usable by unprivileged users */
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	/* the key is a file descriptor */
+	if (attr->max_entries == 0 || attr->key_size != sizeof(int) ||
+	    (attr->map_flags & ~(BPF_F_RDONLY | BPF_F_WRONLY |
+				 BPF_F_RDONLY_PROG)) ||
+	    /* for now, force read-only map for eBPF programs because only
+	     * bpf_inode_map_lookup_elem() enable to access them */
+	    !(attr->map_flags & BPF_F_RDONLY_PROG) ||
+	    bpf_map_attr_numa_node(attr) != NUMA_NO_NODE)
+		return -EINVAL;
+
+	/*
+	 * Limit number of entries in an inode map to the maximum number of
+	 * open files for the current process. The maximum number of file
+	 * references (including all inode maps) for a process is then
+	 * (RLIMIT_NOFILE - 1) * RLIMIT_NOFILE. If the process' RLIMIT_NOFILE
+	 * is 0, then any entry update is forbidden.
+	 *
+	 * An eBPF program can inherit all the inode map FD. The worse case is
+	 * to fill a bunch of arraymaps, create an eBPF program, close the
+	 * inode map FDs, and start again. The maximum number of inode map
+	 * entries can then be close to RLIMIT_NOFILE^3.
+	 */
+	if (attr->max_entries > rlimit(RLIMIT_NOFILE))
+		return -EMFILE;
+
+	/* decorelate UAPI from kernel API */
+	attr->key_size = sizeof(struct inode *);
+
+	return htab_map_alloc_check(attr);
+}
+
+static void inode_htab_put_key(void *key)
+{
+	struct inode **inode = key;
+
+	if ((*inode)->i_state & I_FREEING)
+		return;
+	iput(*inode);
+}
+
+/* called from syscall or (never) from eBPF program */
+static int map_get_next_no_key(struct bpf_map *map, void *key, void *next_key)
+{
+	/* do not leak a file descriptor */
+	return -ENOTSUPP;
+}
+
+/* must call iput(inode) after this call */
+static struct inode *inode_from_fd(int ufd, bool check_access)
+{
+	struct inode *ret;
+	struct fd f;
+	int deny;
+
+	f = fdget(ufd);
+	if (unlikely(!f.file))
+		return ERR_PTR(-EBADF);
+	/* TODO?: add this check when called from an eBPF program too (already
+	* checked by the LSM parent hooks anyway) */
+	if (unlikely(IS_PRIVATE(file_inode(f.file)))) {
+		ret = ERR_PTR(-EINVAL);
+		goto put_fd;
+	}
+	/* check if the FD is tied to a mount point */
+	/* TODO?: add this check when called from an eBPF program too */
+	if (unlikely(f.file->f_path.mnt->mnt_flags & MNT_INTERNAL)) {
+		ret = ERR_PTR(-EINVAL);
+		goto put_fd;
+	}
+	if (check_access) {
+		/*
+		* must be allowed to access attributes from this file to then
+		* be able to compare an inode to its map entry
+		*/
+		deny = security_inode_getattr(&f.file->f_path);
+		if (deny) {
+			ret = ERR_PTR(deny);
+			goto put_fd;
+		}
+	}
+	ret = file_inode(f.file);
+	ihold(ret);
+
+put_fd:
+	fdput(f);
+	return ret;
+}
+
+/*
+ * The key is a FD when called from a syscall, but an inode address when called
+ * from an eBPF program.
+ */
+
+/* called from syscall */
+int bpf_inode_fd_htab_map_lookup_elem(struct bpf_map *map, int *key, void *value)
+{
+	void *ptr;
+	struct inode *inode;
+	int ret;
+
+	/* check inode access */
+	inode = inode_from_fd(*key, true);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+
+	rcu_read_lock();
+	ptr = htab_map_lookup_elem(map, &inode);
+	iput(inode);
+	if (IS_ERR(ptr)) {
+		ret = PTR_ERR(ptr);
+	} else if (!ptr) {
+		ret = -ENOENT;
+	} else {
+		ret = 0;
+		copy_map_value(map, value, ptr);
+	}
+	rcu_read_unlock();
+	return ret;
+}
+
+/* called from kernel */
+int bpf_inode_ptr_locked_htab_map_delete_elem(struct bpf_map *map,
+		struct inode **key, bool remove_in_inode)
+{
+	if (remove_in_inode)
+		landlock_inode_remove_map(*key, map);
+	return htab_map_delete_elem(map, key);
+}
+
+/* called from syscall */
+int bpf_inode_fd_htab_map_delete_elem(struct bpf_map *map, int *key)
+{
+	struct inode *inode;
+	int ret;
+
+	/* do not check inode access (similar to directory check) */
+	inode = inode_from_fd(*key, false);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+	ret = bpf_inode_ptr_locked_htab_map_delete_elem(map, &inode, true);
+	iput(inode);
+	return ret;
+}
+
+/* called from syscall */
+int bpf_inode_fd_htab_map_update_elem(struct bpf_map *map, int *key, void *value,
+		u64 map_flags)
+{
+	struct inode *inode;
+	int ret;
+
+	WARN_ON_ONCE(!rcu_read_lock_held());
+
+	/* check inode access */
+	inode = inode_from_fd(*key, true);
+	if (IS_ERR(inode))
+		return PTR_ERR(inode);
+	ret = htab_map_update_elem(map, &inode, value, map_flags);
+	if (!ret)
+		ret = landlock_inode_add_map(inode, map);
+	iput(inode);
+	return ret;
+}
+
+static void inode_htab_map_free(struct bpf_map *map)
+{
+	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
+	struct hlist_nulls_node *n;
+	struct hlist_nulls_head *head;
+	struct htab_elem *l;
+	int i;
+
+	for (i = 0; i < htab->n_buckets; i++) {
+		head = select_bucket(htab, i);
+		hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
+			landlock_inode_remove_map(*((struct inode **)l->key), map);
+		}
+	}
+	htab_map_free(map);
+}
+
+/* use the map_inode_lookup_elem() helper instead */
+static void *map_lookup_no_elem(struct bpf_map *map, void *key)
+{
+	WARN_ON_ONCE(1);
+	return NULL;
+}
+
+static int map_delete_no_elem(struct bpf_map *map, void *key)
+{
+	WARN_ON_ONCE(1);
+	return -ENOTSUPP;
+}
+
+static int map_update_no_elem(struct bpf_map *map, void *key, void *value,
+		u64 flags)
+{
+	WARN_ON_ONCE(1);
+	return -ENOTSUPP;
+}
+
+const struct bpf_map_ops htab_inode_ops = {
+	.map_alloc_check = inode_htab_map_alloc_check,
+	.map_alloc = htab_map_alloc,
+	.map_free = inode_htab_map_free,
+	.map_put_key = inode_htab_put_key,
+	.map_get_next_key = map_get_next_no_key,
+	.map_lookup_elem = map_lookup_no_elem,
+	.map_delete_elem = map_delete_no_elem,
+	.map_update_elem = map_update_no_elem,
+	.map_check_btf = map_check_no_btf,
+};
+
+/*
+ * We need a dedicated helper to deal with inode maps because the key is a
+ * pointer to an opaque data, only provided by the kernel.  This really act
+ * like a (physical or cryptographic) key, which is why it is also not allowed
+ * to get the next key with map_get_next_key().
+ */
+BPF_CALL_2(bpf_inode_map_lookup_elem, struct bpf_map *, map, void *, key)
+{
+	WARN_ON_ONCE(!rcu_read_lock_held());
+	return (unsigned long)htab_map_lookup_elem(map, &key);
+}
+
+const struct bpf_func_proto bpf_inode_map_lookup_elem_proto = {
+	.func		= bpf_inode_map_lookup_elem,
+	.gpl_only	= false,
+	.pkt_access	= true,
+	.ret_type	= RET_PTR_TO_MAP_VALUE_OR_NULL,
+	.arg1_type	= ARG_CONST_MAP_PTR,
+	.arg2_type	= ARG_PTR_TO_INODE,
+};
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index b2a8cb14f28e..e46441c42b68 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -801,6 +801,8 @@ static int map_lookup_elem(union bpf_attr *attr)
 	} else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
 		   map->map_type == BPF_MAP_TYPE_STACK) {
 		err = map->ops->map_peek_elem(map, value);
+	} else if (map->map_type == BPF_MAP_TYPE_INODE) {
+		err = bpf_inode_fd_htab_map_lookup_elem(map, key, value);
 	} else {
 		rcu_read_lock();
 		if (map->ops->map_lookup_elem_sys_only)
@@ -951,6 +953,10 @@ static int map_update_elem(union bpf_attr *attr)
 	} else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
 		   map->map_type == BPF_MAP_TYPE_STACK) {
 		err = map->ops->map_push_elem(map, value, attr->flags);
+	} else if (map->map_type == BPF_MAP_TYPE_INODE) {
+		rcu_read_lock();
+		err = bpf_inode_fd_htab_map_update_elem(map, key, value, attr->flags);
+		rcu_read_unlock();
 	} else {
 		rcu_read_lock();
 		err = map->ops->map_update_elem(map, key, value, attr->flags);
@@ -1006,7 +1012,10 @@ static int map_delete_elem(union bpf_attr *attr)
 	preempt_disable();
 	__this_cpu_inc(bpf_prog_active);
 	rcu_read_lock();
-	err = map->ops->map_delete_elem(map, key);
+	if (map->map_type == BPF_MAP_TYPE_INODE)
+		err = bpf_inode_fd_htab_map_delete_elem(map, key);
+	else
+		err = map->ops->map_delete_elem(map, key);
 	rcu_read_unlock();
 	__this_cpu_dec(bpf_prog_active);
 	preempt_enable();
@@ -1018,6 +1027,22 @@ static int map_delete_elem(union bpf_attr *attr)
 	return err;
 }
 
+int bpf_inode_ptr_unlocked_htab_map_delete_elem(struct bpf_map *map,
+						struct inode **key, bool remove_in_inode)
+{
+	int err;
+
+	preempt_disable();
+	__this_cpu_inc(bpf_prog_active);
+	rcu_read_lock();
+	err = bpf_inode_ptr_locked_htab_map_delete_elem(map, key, remove_in_inode);
+	rcu_read_unlock();
+	__this_cpu_dec(bpf_prog_active);
+	preempt_enable();
+	maybe_wait_bpf_programs(map);
+	return err;
+}
+
 /* last field in 'union bpf_attr' used by this command */
 #define BPF_MAP_GET_NEXT_KEY_LAST_FIELD next_key
 
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 026c68cb9116..3972b9f02dac 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -400,6 +400,7 @@ static const char * const reg_type_str[] = {
 	[PTR_TO_TCP_SOCK_OR_NULL] = "tcp_sock_or_null",
 	[PTR_TO_TP_BUFFER]	= "tp_buffer",
 	[PTR_TO_XDP_SOCK]	= "xdp_sock",
+	[PTR_TO_INODE]		= "inode",
 };
 
 static char slot_type_char[] = {
@@ -1846,6 +1847,7 @@ static bool is_spillable_regtype(enum bpf_reg_type type)
 	case PTR_TO_TCP_SOCK:
 	case PTR_TO_TCP_SOCK_OR_NULL:
 	case PTR_TO_XDP_SOCK:
+	case PTR_TO_INODE:
 		return true;
 	default:
 		return false;
@@ -3306,6 +3308,10 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 			verbose(env, "verifier internal error\n");
 			return -EFAULT;
 		}
+	} else if (arg_type == ARG_PTR_TO_INODE) {
+		expected_type = PTR_TO_INODE;
+		if (type != expected_type)
+			goto err_type;
 	} else if (arg_type_is_mem_ptr(arg_type)) {
 		expected_type = PTR_TO_STACK;
 		/* One exception here. In case function allows for NULL to be
@@ -3511,6 +3517,10 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env,
 		    func_id != BPF_FUNC_sk_storage_delete)
 			goto error;
 		break;
+	case BPF_MAP_TYPE_INODE:
+		if (func_id != BPF_FUNC_inode_map_lookup_elem)
+			goto error;
+		break;
 	default:
 		break;
 	}
@@ -3579,6 +3589,10 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env,
 		if (map->map_type != BPF_MAP_TYPE_SK_STORAGE)
 			goto error;
 		break;
+	case BPF_FUNC_inode_map_lookup_elem:
+		if (map->map_type != BPF_MAP_TYPE_INODE)
+			goto error;
+		break;
 	default:
 		break;
 	}
diff --git a/security/landlock/common.h b/security/landlock/common.h
index b2ee018eb6fc..b0ba3f31ac7d 100644
--- a/security/landlock/common.h
+++ b/security/landlock/common.h
@@ -11,6 +11,7 @@
 
 #include <linux/bpf.h> /* enum bpf_attach_type */
 #include <linux/filter.h> /* bpf_prog */
+#include <linux/lsm_hooks.h> /* lsm_blob_sizes */
 #include <linux/refcount.h> /* refcount_t */
 #include <uapi/linux/landlock.h> /* LANDLOCK_TRIGGER_* */
 
@@ -23,6 +24,8 @@
 #define _LANDLOCK_TRIGGER_FS_PICK_LAST	LANDLOCK_TRIGGER_FS_PICK_WRITE
 #define _LANDLOCK_TRIGGER_FS_PICK_MASK	((_LANDLOCK_TRIGGER_FS_PICK_LAST << 1ULL) - 1)
 
+extern struct lsm_blob_sizes landlock_blob_sizes;
+
 enum landlock_hook_type {
 	LANDLOCK_HOOK_FS_PICK = 1,
 	LANDLOCK_HOOK_FS_WALK,
@@ -55,6 +58,17 @@ struct landlock_prog_set {
 	refcount_t usage;
 };
 
+struct landlock_inode_map {
+	struct list_head list;
+	struct rcu_head rcu_put;
+	struct bpf_map *map;
+	/*
+	 * It would be nice to remove the inode field, but it is necessary for
+	 * call_rcu() .
+	 */
+	struct inode *inode;
+};
+
 /**
  * get_hook_index - get an index for the programs of struct landlock_prog_set
  *
diff --git a/security/landlock/hooks_fs.c b/security/landlock/hooks_fs.c
index 3f81b7fc2938..8c9d6a333111 100644
--- a/security/landlock/hooks_fs.c
+++ b/security/landlock/hooks_fs.c
@@ -46,6 +46,12 @@ bool landlock_is_valid_access_fs_pick(int off, enum bpf_access_type type,
 		enum bpf_reg_type *reg_type, int *max_size)
 {
 	switch (off) {
+	case offsetof(struct landlock_ctx_fs_pick, inode):
+		if (type != BPF_READ)
+			return false;
+		*reg_type = PTR_TO_INODE;
+		*max_size = sizeof(u64);
+		return true;
 	default:
 		return false;
 	}
@@ -55,6 +61,12 @@ bool landlock_is_valid_access_fs_walk(int off, enum bpf_access_type type,
 		enum bpf_reg_type *reg_type, int *max_size)
 {
 	switch (off) {
+	case offsetof(struct landlock_ctx_fs_walk, inode):
+		if (type != BPF_READ)
+			return false;
+		*reg_type = PTR_TO_INODE;
+		*max_size = sizeof(u64);
+		return true;
 	default:
 		return false;
 	}
@@ -237,8 +249,79 @@ static int hook_sb_pivotroot(const struct path *old_path,
 			new_path->dentry->d_inode);
 }
 
+/* inode helpers */
+
+static inline struct list_head *inode_landlock(const struct inode *inode)
+{
+	return inode->i_security + landlock_blob_sizes.lbs_inode;
+}
+
+int landlock_inode_add_map(struct inode *inode, struct bpf_map *map)
+{
+	struct landlock_inode_map *inode_map;
+
+	inode_map = kzalloc(sizeof(*inode_map), GFP_ATOMIC);
+	if (!inode_map)
+		return -ENOMEM;
+	INIT_LIST_HEAD(&inode_map->list);
+	inode_map->map = map;
+	inode_map->inode = inode;
+	list_add_tail(&inode_map->list, inode_landlock(inode));
+	return 0;
+}
+
+static void put_landlock_inode_map(struct rcu_head *head)
+{
+	struct landlock_inode_map *inode_map;
+	int err;
+
+	inode_map = container_of(head, struct landlock_inode_map, rcu_put);
+	err = bpf_inode_ptr_unlocked_htab_map_delete_elem(inode_map->map,
+			&inode_map->inode, false);
+	bpf_map_put(inode_map->map);
+	kfree(inode_map);
+}
+
+void landlock_inode_remove_map(struct inode *inode, const struct bpf_map *map)
+{
+	struct landlock_inode_map *inode_map;
+	bool found = false;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(inode_map, inode_landlock(inode), list) {
+		if (inode_map->map == map) {
+			found = true;
+			list_del_rcu(&inode_map->list);
+			kfree_rcu(inode_map, rcu_put);
+			break;
+		}
+	}
+	rcu_read_unlock();
+	WARN_ON(!found);
+}
+
 /* inode hooks */
 
+static int hook_inode_alloc_security(struct inode *inode)
+{
+	struct list_head *ll_inode = inode_landlock(inode);
+
+	INIT_LIST_HEAD(ll_inode);
+	return 0;
+}
+
+static void hook_inode_free_security(struct inode *inode)
+{
+	struct landlock_inode_map *inode_map;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(inode_map, inode_landlock(inode), list) {
+		list_del_rcu(&inode_map->list);
+		call_rcu(&inode_map->rcu_put, put_landlock_inode_map);
+	}
+	rcu_read_unlock();
+}
+
 /* a directory inode contains only one dentry */
 static int hook_inode_create(struct inode *dir, struct dentry *dentry,
 		umode_t mode)
@@ -517,6 +600,8 @@ static struct security_hook_list landlock_hooks[] = {
 	LSM_HOOK_INIT(sb_mount, hook_sb_mount),
 	LSM_HOOK_INIT(sb_pivotroot, hook_sb_pivotroot),
 
+	LSM_HOOK_INIT(inode_alloc_security, hook_inode_alloc_security),
+	LSM_HOOK_INIT(inode_free_security, hook_inode_free_security),
 	LSM_HOOK_INIT(inode_create, hook_inode_create),
 	LSM_HOOK_INIT(inode_link, hook_inode_link),
 	LSM_HOOK_INIT(inode_unlink, hook_inode_unlink),
diff --git a/security/landlock/init.c b/security/landlock/init.c
index 391e88bd4d3a..eec4467cb5ee 100644
--- a/security/landlock/init.c
+++ b/security/landlock/init.c
@@ -104,6 +104,18 @@ static const struct bpf_func_proto *bpf_landlock_func_proto(
 	default:
 		break;
 	}
+
+	switch (get_hook_type(prog)) {
+	case LANDLOCK_HOOK_FS_WALK:
+	case LANDLOCK_HOOK_FS_PICK:
+		switch (func_id) {
+		case BPF_FUNC_inode_map_lookup_elem:
+			return &bpf_inode_map_lookup_elem_proto;
+		default:
+			break;
+		}
+		break;
+	}
 	return NULL;
 }
 
@@ -123,6 +135,7 @@ static int __init landlock_init(void)
 }
 
 struct lsm_blob_sizes landlock_blob_sizes __lsm_ro_after_init = {
+	.lbs_inode = sizeof(struct list_head),
 };
 
 DEFINE_LSM(LANDLOCK_NAME) = {
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index 7b7a4f6c3104..7a55535f5dc1 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -134,6 +134,7 @@ enum bpf_map_type {
 	BPF_MAP_TYPE_QUEUE,
 	BPF_MAP_TYPE_STACK,
 	BPF_MAP_TYPE_SK_STORAGE,
+	BPF_MAP_TYPE_INODE,
 };
 
 /* Note that tracing related programs such as
@@ -2714,6 +2715,14 @@ union bpf_attr {
  *		**-EPERM** if no permission to send the *sig*.
  *
  *		**-EAGAIN** if bpf program can try again.
+ *
+ * void *bpf_inode_map_lookup_elem(struct bpf_map *map, const void *key)
+ *	Description
+ *		Perform a lookup in *map* for an entry associated to an inode
+ *		*key*.
+ *	Return
+ *		Map value associated to *key*, or **NULL** if no entry was
+ *		found.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
@@ -2825,7 +2834,8 @@ union bpf_attr {
 	FN(strtoul),			\
 	FN(sk_storage_get),		\
 	FN(sk_storage_delete),		\
-	FN(send_signal),
+	FN(send_signal),		\
+	FN(inode_map_lookup_elem),
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
  * function eBPF program intends to call
diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c
index 03c910d1f84c..98875221310d 100644
--- a/tools/lib/bpf/libbpf_probes.c
+++ b/tools/lib/bpf/libbpf_probes.c
@@ -250,6 +250,7 @@ bool bpf_probe_map_type(enum bpf_map_type map_type, __u32 ifindex)
 	case BPF_MAP_TYPE_XSKMAP:
 	case BPF_MAP_TYPE_SOCKHASH:
 	case BPF_MAP_TYPE_REUSEPORT_SOCKARRAY:
+	case BPF_MAP_TYPE_INODE:
 	default:
 		break;
 	}
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 07/10] landlock: Add ptrace restrictions
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
                   ` (5 preceding siblings ...)
  2019-07-21 21:31 ` [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 08/10] bpf: Add a Landlock sandbox example Mickaël Salaün
                   ` (2 subsequent siblings)
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

A landlocked process has less privileges than a non-landlocked process
and must then be subject to additional restrictions when manipulating
processes. To be allowed to use ptrace(2) and related syscalls on a
target process, a landlocked process must have a subset of the target
process' rules.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: James Morris <jmorris@namei.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
---

Changes since v6:
* factor out ptrace check
* constify pointers
* cleanup headers
* use the new security_add_hooks()
---
 security/landlock/Makefile       |   2 +-
 security/landlock/hooks_ptrace.c | 121 +++++++++++++++++++++++++++++++
 security/landlock/hooks_ptrace.h |   8 ++
 security/landlock/init.c         |   2 +
 4 files changed, 132 insertions(+), 1 deletion(-)
 create mode 100644 security/landlock/hooks_ptrace.c
 create mode 100644 security/landlock/hooks_ptrace.h

diff --git a/security/landlock/Makefile b/security/landlock/Makefile
index 270ece5d93de..4500ddb0767e 100644
--- a/security/landlock/Makefile
+++ b/security/landlock/Makefile
@@ -2,4 +2,4 @@ obj-$(CONFIG_SECURITY_LANDLOCK) := landlock.o
 
 landlock-y := init.o \
 	enforce.o enforce_seccomp.o \
-	hooks.o hooks_fs.o
+	hooks.o hooks_fs.o hooks_ptrace.o
diff --git a/security/landlock/hooks_ptrace.c b/security/landlock/hooks_ptrace.c
new file mode 100644
index 000000000000..7f5e8b994e93
--- /dev/null
+++ b/security/landlock/hooks_ptrace.c
@@ -0,0 +1,121 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Landlock LSM - ptrace hooks
+ *
+ * Copyright © 2017 Mickaël Salaün <mic@digikod.net>
+ */
+
+#include <asm/current.h>
+#include <linux/errno.h>
+#include <linux/kernel.h> /* ARRAY_SIZE */
+#include <linux/lsm_hooks.h>
+#include <linux/sched.h> /* struct task_struct */
+#include <linux/seccomp.h>
+
+#include "common.h" /* struct landlock_prog_set */
+#include "hooks.h" /* landlocked() */
+#include "hooks_ptrace.h"
+
+static bool progs_are_subset(const struct landlock_prog_set *parent,
+		const struct landlock_prog_set *child)
+{
+	size_t i;
+
+	if (!parent || !child)
+		return false;
+	if (parent == child)
+		return true;
+
+	for (i = 0; i < ARRAY_SIZE(child->programs); i++) {
+		struct landlock_prog_list *walker;
+		bool found_parent = false;
+
+		if (!parent->programs[i])
+			continue;
+		for (walker = child->programs[i]; walker;
+				walker = walker->prev) {
+			if (walker == parent->programs[i]) {
+				found_parent = true;
+				break;
+			}
+		}
+		if (!found_parent)
+			return false;
+	}
+	return true;
+}
+
+static bool task_has_subset_progs(const struct task_struct *parent,
+		const struct task_struct *child)
+{
+#ifdef CONFIG_SECCOMP_FILTER
+	if (progs_are_subset(parent->seccomp.landlock_prog_set,
+				child->seccomp.landlock_prog_set))
+		/* must be ANDed with other providers (i.e. cgroup) */
+		return true;
+#endif /* CONFIG_SECCOMP_FILTER */
+	return false;
+}
+
+static int task_ptrace(const struct task_struct *parent,
+		const struct task_struct *child)
+{
+	if (!landlocked(parent))
+		return 0;
+
+	if (!landlocked(child))
+		return -EPERM;
+
+	if (task_has_subset_progs(parent, child))
+		return 0;
+
+	return -EPERM;
+}
+
+/**
+ * hook_ptrace_access_check - determine whether the current process may access
+ *			      another
+ *
+ * @child: the process to be accessed
+ * @mode: the mode of attachment
+ *
+ * If the current task has Landlock programs, then the child must have at least
+ * the same programs.  Else denied.
+ *
+ * Determine whether a process may access another, returning 0 if permission
+ * granted, -errno if denied.
+ */
+static int hook_ptrace_access_check(struct task_struct *child,
+		unsigned int mode)
+{
+	return task_ptrace(current, child);
+}
+
+/**
+ * hook_ptrace_traceme - determine whether another process may trace the
+ *			 current one
+ *
+ * @parent: the task proposed to be the tracer
+ *
+ * If the parent has Landlock programs, then the current task must have the
+ * same or more programs.
+ * Else denied.
+ *
+ * Determine whether the nominated task is permitted to trace the current
+ * process, returning 0 if permission is granted, -errno if denied.
+ */
+static int hook_ptrace_traceme(struct task_struct *parent)
+{
+	return task_ptrace(parent, current);
+}
+
+static struct security_hook_list landlock_hooks[] = {
+	LSM_HOOK_INIT(ptrace_access_check, hook_ptrace_access_check),
+	LSM_HOOK_INIT(ptrace_traceme, hook_ptrace_traceme),
+};
+
+__init void landlock_add_hooks_ptrace(void)
+{
+	security_add_hooks(landlock_hooks, ARRAY_SIZE(landlock_hooks),
+			LANDLOCK_NAME);
+}
diff --git a/security/landlock/hooks_ptrace.h b/security/landlock/hooks_ptrace.h
new file mode 100644
index 000000000000..2c2b8a13037f
--- /dev/null
+++ b/security/landlock/hooks_ptrace.h
@@ -0,0 +1,8 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Landlock LSM - ptrace hooks
+ *
+ * Copyright © 2017 Mickaël Salaün <mic@digikod.net>
+ */
+
+__init void landlock_add_hooks_ptrace(void);
diff --git a/security/landlock/init.c b/security/landlock/init.c
index eec4467cb5ee..35165fc8a595 100644
--- a/security/landlock/init.c
+++ b/security/landlock/init.c
@@ -13,6 +13,7 @@
 
 #include "common.h" /* LANDLOCK_* */
 #include "hooks_fs.h"
+#include "hooks_ptrace.h"
 
 static bool bpf_landlock_is_valid_access(int off, int size,
 		enum bpf_access_type type, const struct bpf_prog *prog,
@@ -130,6 +131,7 @@ const struct bpf_prog_ops landlock_prog_ops = {};
 static int __init landlock_init(void)
 {
 	pr_info(LANDLOCK_NAME ": Initializing (sandbox with seccomp)\n");
+	landlock_add_hooks_ptrace();
 	landlock_add_hooks_fs();
 	return 0;
 }
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 08/10] bpf: Add a Landlock sandbox example
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
                   ` (6 preceding siblings ...)
  2019-07-21 21:31 ` [PATCH bpf-next v10 07/10] landlock: Add ptrace restrictions Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 09/10] bpf,landlock: Add tests for Landlock Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation " Mickaël Salaün
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

Add a basic sandbox tool to launch a command which is denied access to a
list of files and directories.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: James Morris <jmorris@namei.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
---

Changes since v9:
* replace subtype with expected_attach_type and expected_attach_triggers
* add the ability to parse Landlock programs and triggers to libbpf
* use the new bpf_inode_map_lookup_elem()
* use read-only inode map for Landlock programs
* remove bpf_load.c modifications

Changes since v8:
* rewrite the landlock1 sample which deny access to a set of files or
  directories (i.e. simple blacklist) to fit with the previous patches
* add "landlock1" to .gitignore
* in bpf_load.c, pass the subtype with a call to
  bpf_load_program_xattr()

Changes since v7:
* rewrite the example using an inode map
* add to bpf_load the ability to handle subtypes per program type

Changes since v6:
* check return value of load_and_attach()
* allow to write on pipes
* rename BPF_PROG_TYPE_LANDLOCK to BPF_PROG_TYPE_LANDLOCK_RULE
* rename Landlock version to ABI to better reflect its purpose
* use const variable (suggested by Kees Cook)
* remove useless definitions (suggested by Kees Cook)
* add detailed explanations (suggested by Kees Cook)

Changes since v5:
* cosmetic fixes
* rebase

Changes since v4:
* write Landlock rule in C and compiled it with LLVM
* remove cgroup handling
* remove path handling: only handle a read-only environment
* remove errno return codes

Changes since v3:
* remove seccomp and origin field: completely free from seccomp programs
* handle more FS-related hooks
* handle inode hooks and directory traversal
* add faked but consistent view thanks to ENOENT
* add /lib64 in the example
* fix spelling
* rename some types and definitions (e.g. SECCOMP_ADD_LANDLOCK_RULE)

Changes since v2:
* use BPF_PROG_ATTACH for cgroup handling
---
 samples/bpf/.gitignore                        |   1 +
 samples/bpf/Makefile                          |   3 +
 samples/bpf/landlock1.h                       |   8 +
 samples/bpf/landlock1_kern.c                  |  55 ++++
 samples/bpf/landlock1_user.c                  | 250 ++++++++++++++++++
 tools/lib/bpf/libbpf.c                        |  43 ++-
 tools/lib/bpf/libbpf.h                        |   7 +-
 tools/lib/bpf/libbpf.map                      |   1 +
 tools/testing/selftests/bpf/bpf_helpers.h     |   2 +
 .../selftests/bpf/test_section_names.c        |   2 +-
 .../selftests/bpf/test_sockopt_multi.c        |   4 +-
 tools/testing/selftests/bpf/test_sockopt_sk.c |   2 +-
 12 files changed, 364 insertions(+), 14 deletions(-)
 create mode 100644 samples/bpf/landlock1.h
 create mode 100644 samples/bpf/landlock1_kern.c
 create mode 100644 samples/bpf/landlock1_user.c

diff --git a/samples/bpf/.gitignore b/samples/bpf/.gitignore
index 74d31fd3c99c..a4c9c806f739 100644
--- a/samples/bpf/.gitignore
+++ b/samples/bpf/.gitignore
@@ -2,6 +2,7 @@ cpustat
 fds_example
 hbm
 ibumad
+landlock1
 lathist
 lwt_len_hist
 map_perf_test
diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
index f90daadfbc89..b0309ed7c1c9 100644
--- a/samples/bpf/Makefile
+++ b/samples/bpf/Makefile
@@ -53,6 +53,7 @@ hostprogs-y += task_fd_query
 hostprogs-y += xdp_sample_pkts
 hostprogs-y += ibumad
 hostprogs-y += hbm
+hostprogs-y += landlock1
 
 # Libbpf dependencies
 LIBBPF = $(TOOLS_PATH)/lib/bpf/libbpf.a
@@ -109,6 +110,7 @@ task_fd_query-objs := bpf_load.o task_fd_query_user.o $(TRACE_HELPERS)
 xdp_sample_pkts-objs := xdp_sample_pkts_user.o $(TRACE_HELPERS)
 ibumad-objs := bpf_load.o ibumad_user.o $(TRACE_HELPERS)
 hbm-objs := bpf_load.o hbm.o $(CGROUP_HELPERS)
+landlock1-objs := bpf_load.o landlock1_user.o
 
 # Tell kbuild to always build the programs
 always := $(hostprogs-y)
@@ -170,6 +172,7 @@ always += xdp_sample_pkts_kern.o
 always += ibumad_kern.o
 always += hbm_out_kern.o
 always += hbm_edt_kern.o
+always += landlock1_kern.o
 
 KBUILD_HOSTCFLAGS += -I$(objtree)/usr/include
 KBUILD_HOSTCFLAGS += -I$(srctree)/tools/lib/bpf/
diff --git a/samples/bpf/landlock1.h b/samples/bpf/landlock1.h
new file mode 100644
index 000000000000..53b0a9447855
--- /dev/null
+++ b/samples/bpf/landlock1.h
@@ -0,0 +1,8 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Landlock sample 1 - common header
+ *
+ * Copyright © 2018-2019 Mickaël Salaün <mic@digikod.net>
+ */
+
+#define MAP_FLAG_DENY		(1ULL << 0)
diff --git a/samples/bpf/landlock1_kern.c b/samples/bpf/landlock1_kern.c
new file mode 100644
index 000000000000..d6946659f891
--- /dev/null
+++ b/samples/bpf/landlock1_kern.c
@@ -0,0 +1,55 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Landlock sample 1 - whitelist of read only or read-write file hierarchy
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ */
+
+/*
+ * This file contains a function that will be compiled to eBPF bytecode thanks
+ * to LLVM/Clang.
+ *
+ * Each SEC() means that the following function or variable will be part of a
+ * custom ELF section. This sections are then processed by the userspace part
+ * (see landlock1_user.c) to extract eBPF bytecode and metadata.
+ */
+
+#include <uapi/linux/bpf.h>
+#include <uapi/linux/landlock.h>
+
+#include "bpf_helpers.h"
+#include "landlock1.h" /* MAP_FLAG_DENY */
+
+#define MAP_MAX_ENTRIES		20
+
+struct bpf_map_def SEC("maps") inode_map = {
+	.type = BPF_MAP_TYPE_INODE,
+	.key_size = sizeof(u32),
+	.value_size = sizeof(u64),
+	.max_entries = MAP_MAX_ENTRIES,
+	.map_flags = BPF_F_RDONLY_PROG,
+};
+
+static __always_inline __u64 get_access(void *inode)
+{
+	u64 *flags;
+
+	flags = bpf_inode_map_lookup_elem(&inode_map, inode);
+	if (flags && (*flags & MAP_FLAG_DENY))
+		return LANDLOCK_RET_DENY;
+	return LANDLOCK_RET_ALLOW;
+}
+
+SEC("landlock/fs_walk")
+int fs_walk(struct landlock_ctx_fs_walk *ctx)
+{
+	return get_access((void *)ctx->inode);
+}
+
+SEC("landlock/fs_pick")
+int fs_pick_ro(struct landlock_ctx_fs_pick *ctx)
+{
+	return get_access((void *)ctx->inode);
+}
+
+static const char SEC("license") _license[] = "GPL";
diff --git a/samples/bpf/landlock1_user.c b/samples/bpf/landlock1_user.c
new file mode 100644
index 000000000000..2082ca367f94
--- /dev/null
+++ b/samples/bpf/landlock1_user.c
@@ -0,0 +1,250 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Landlock sample 1 - deny access to a set of directories (blacklisting)
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ */
+
+#include "bpf/libbpf.h"
+#include "bpf_load.h"
+#include "landlock1.h" /* MAP_FLAG_DENY */
+
+#define _GNU_SOURCE
+#include <errno.h>
+#include <fcntl.h> /* open() */
+#include <linux/bpf.h>
+#include <linux/filter.h>
+#include <linux/landlock.h>
+#include <linux/prctl.h>
+#include <linux/seccomp.h>
+#include <stddef.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/prctl.h>
+#include <sys/syscall.h>
+#include <unistd.h>
+
+#ifndef seccomp
+static int seccomp(unsigned int op, unsigned int flags, void *args)
+{
+	errno = 0;
+	return syscall(__NR_seccomp, op, flags, args);
+}
+#endif
+
+static int apply_sandbox(int prog_fd)
+{
+	int ret = 0;
+
+	/* set up the test sandbox */
+	if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) {
+		perror("prctl(no_new_priv)");
+		return 1;
+	}
+	if (seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, &prog_fd)) {
+		perror("seccomp(set_hook)");
+		ret = 1;
+	}
+	close(prog_fd);
+
+	return ret;
+}
+
+#define ENV_FS_PATH_DENY_NAME "LL_PATH_DENY"
+#define ENV_PATH_TOKEN ":"
+
+static int parse_path(char *env_path, const char ***path_list)
+{
+	int i, path_nb = 0;
+
+	if (env_path) {
+		path_nb++;
+		for (i = 0; env_path[i]; i++) {
+			if (env_path[i] == ENV_PATH_TOKEN[0])
+				path_nb++;
+		}
+	}
+	*path_list = malloc(path_nb * sizeof(**path_list));
+	for (i = 0; i < path_nb; i++)
+		(*path_list)[i] = strsep(&env_path, ENV_PATH_TOKEN);
+
+	return path_nb;
+}
+
+static int populate_map(const char *env_var, unsigned long long value,
+		int map_fd)
+{
+	int path_nb, ref_fd, i;
+	char *env_path_name;
+	const char **path_list = NULL;
+
+	env_path_name = getenv(env_var);
+	if (!env_path_name)
+		return 0;
+	env_path_name = strdup(env_path_name);
+	path_nb = parse_path(env_path_name, &path_list);
+
+	for (i = 0; i < path_nb; i++) {
+		ref_fd = open(path_list[i], O_RDONLY | O_CLOEXEC);
+		if (ref_fd < 0) {
+			fprintf(stderr, "Failed to open \"%s\": %s\n",
+					path_list[i],
+					strerror(errno));
+			return 1;
+		}
+		if (bpf_map_update_elem(map_fd, &ref_fd, &value, BPF_ANY)) {
+			fprintf(stderr, "Failed to update the map with"
+					" \"%s\": %s\n", path_list[i],
+					strerror(errno));
+			return 1;
+		}
+		close(ref_fd);
+	}
+	free(env_path_name);
+	return 0;
+}
+
+/* need to call bpf_object__close(obj) once every FD is used */
+static int ll_load_file(const char *filename, struct bpf_object **obj,
+		int *ll_map, int *ll_prog_walk, int *ll_prog_pick)
+{
+	int first_bpf_prog, map_fd, prog_walk_fd, prog_pick_fd, err;
+	struct bpf_map *map;
+	struct bpf_program *prog;
+	struct bpf_object *tmp_obj;
+	struct bpf_prog_load_attr prog_load_attr = {
+		.prog_type = BPF_PROG_TYPE_UNSPEC,
+		.file = filename,
+	};
+
+	/*
+	 * allowed:
+	 * - LANDLOCK_TRIGGER_FS_PICK_LINK
+	 * - LANDLOCK_TRIGGER_FS_PICK_LINKTO
+	 * - LANDLOCK_TRIGGER_FS_PICK_RECEIVE
+	 * - LANDLOCK_TRIGGER_FS_PICK_MOUNTON
+	 */
+	prog_load_attr.expected_attach_triggers =
+		LANDLOCK_TRIGGER_FS_PICK_APPEND |
+		LANDLOCK_TRIGGER_FS_PICK_CHDIR |
+		LANDLOCK_TRIGGER_FS_PICK_CHROOT |
+		LANDLOCK_TRIGGER_FS_PICK_CREATE |
+		LANDLOCK_TRIGGER_FS_PICK_EXECUTE |
+		LANDLOCK_TRIGGER_FS_PICK_FCNTL |
+		LANDLOCK_TRIGGER_FS_PICK_GETATTR |
+		LANDLOCK_TRIGGER_FS_PICK_IOCTL |
+		LANDLOCK_TRIGGER_FS_PICK_LOCK |
+		LANDLOCK_TRIGGER_FS_PICK_MAP |
+		LANDLOCK_TRIGGER_FS_PICK_OPEN |
+		LANDLOCK_TRIGGER_FS_PICK_READ |
+		LANDLOCK_TRIGGER_FS_PICK_READDIR |
+		LANDLOCK_TRIGGER_FS_PICK_RENAME |
+		LANDLOCK_TRIGGER_FS_PICK_RENAMETO |
+		LANDLOCK_TRIGGER_FS_PICK_RMDIR |
+		LANDLOCK_TRIGGER_FS_PICK_SETATTR |
+		LANDLOCK_TRIGGER_FS_PICK_TRANSFER |
+		LANDLOCK_TRIGGER_FS_PICK_UNLINK |
+		LANDLOCK_TRIGGER_FS_PICK_WRITE;
+
+	if (access(filename, O_RDONLY) < 0) {
+		printf("Failed to access file %s: %s\n", filename,
+				strerror(errno));
+		return 1;
+	}
+	err = bpf_prog_load_xattr(&prog_load_attr, &tmp_obj, &first_bpf_prog);
+	if (err) {
+		printf("Failed to parse file %s: %s\n", filename, strerror(err));
+		goto error_load;
+	}
+
+	map = bpf_object__find_map_by_name(tmp_obj, "inode_map");
+	map_fd = bpf_map__fd(map);
+	if (map_fd < 0) {
+		printf("Map not found: %s\n", strerror(map_fd));
+		goto put_obj;
+	}
+
+	prog = bpf_object__find_program_by_title(tmp_obj, "landlock/fs_walk");
+	if (!prog) {
+		printf("Program for FS_WALK not found in file %s\n", filename);
+		goto put_obj;
+	}
+	prog_walk_fd = bpf_program__fd(prog);
+	if (prog_walk_fd < 0) {
+		printf("Failed to load the FS_WALK program from file %s\n",
+				strerror(prog_walk_fd));
+		goto put_obj;
+	}
+
+	prog = bpf_object__find_program_by_title(tmp_obj, "landlock/fs_pick");
+	if (!prog) {
+		printf("Failed to get a file descriptor for program %s from file %s\n",
+				bpf_program__title(prog, false), filename);
+		goto put_obj;
+	}
+	prog_pick_fd = bpf_program__fd(prog);
+	if (prog_pick_fd < 0) {
+		printf("Failed to get a file descriptor for program %s from file %s\n",
+				bpf_program__title(prog, false), filename);
+		goto put_obj;
+	}
+
+	*obj = tmp_obj;
+	*ll_prog_walk = prog_walk_fd;
+	*ll_prog_pick = prog_pick_fd;
+	*ll_map = map_fd;
+	return 0;
+
+put_obj:
+	/* All FDs are closed with bpf_object__close() */
+	bpf_object__close(tmp_obj);
+error_load:
+	printf("ERROR: load_bpf_file failed for: %s\n", filename);
+	printf("  Output from verifier:\n%s\n------\n", bpf_log_buf);
+	return 1;
+}
+
+int main(int argc, char * const argv[], char * const *envp)
+{
+	char filename[256];
+	char *cmd_path;
+	char * const *cmd_argv;
+	struct bpf_object *obj;
+	int ll_map, ll_prog_walk, ll_prog_pick;
+
+	if (argc < 2) {
+		fprintf(stderr, "usage: %s <cmd> [args]...\n\n", argv[0]);
+		fprintf(stderr, "Launch a command in a restricted environment.\n\n");
+		fprintf(stderr, "Environment variables containing paths, each separated by a colon:\n");
+		fprintf(stderr, "* %s: list of files and directories which are denied\n",
+				ENV_FS_PATH_DENY_NAME);
+		fprintf(stderr, "\nexample:\n"
+				"%s=\"${HOME}/.ssh:${HOME}/Images\" "
+				"%s /bin/sh -i\n",
+				ENV_FS_PATH_DENY_NAME, argv[0]);
+		return 1;
+	}
+
+	snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
+	if (ll_load_file(filename, &obj, &ll_map, &ll_prog_walk, &ll_prog_pick))
+		return 1;
+
+	if (populate_map(ENV_FS_PATH_DENY_NAME, MAP_FLAG_DENY, ll_map))
+		return 1;
+	//close(ll_map);
+
+	fprintf(stderr, "Launching a new sandboxed process\n");
+	if (apply_sandbox(ll_prog_walk))
+		return 1;
+	//close(ll_prog_walk);
+	if (apply_sandbox(ll_prog_pick))
+		return 1;
+	//close(ll_prog_pick);
+	//bpf_object__close(obj);
+	cmd_path = argv[1];
+	cmd_argv = argv + 1;
+	execve(cmd_path, cmd_argv, envp);
+	perror("Failed to call execve");
+	return 1;
+}
diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index ab3b8b510b8a..f043e97bca0c 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -181,6 +181,7 @@ struct bpf_program {
 	bpf_program_clear_priv_t clear_priv;
 
 	enum bpf_attach_type expected_attach_type;
+	__u64 expected_attach_triggers;
 	int btf_fd;
 	void *func_info;
 	__u32 func_info_rec_size;
@@ -2459,6 +2460,7 @@ load_program(struct bpf_program *prog, struct bpf_insn *insns, int insns_cnt,
 	memset(&load_attr, 0, sizeof(struct bpf_load_program_attr));
 	load_attr.prog_type = prog->type;
 	load_attr.expected_attach_type = prog->expected_attach_type;
+	load_attr.expected_attach_triggers = prog->expected_attach_triggers;
 	if (prog->caps->name)
 		load_attr.name = prog->name;
 	load_attr.insns = insns;
@@ -3540,19 +3542,29 @@ void bpf_program__set_expected_attach_type(struct bpf_program *prog,
 	prog->expected_attach_type = type;
 }
 
-#define BPF_PROG_SEC_IMPL(string, ptype, eatype, is_attachable, atype) \
-	{ string, sizeof(string) - 1, ptype, eatype, is_attachable, atype }
+void bpf_program__set_expected_attach_triggers(struct bpf_program *prog,
+					       __u64 triggers)
+{
+	prog->expected_attach_triggers = triggers;
+}
+
+#define BPF_PROG_SEC_IMPL(string, ptype, eatype, is_attachable, atype, has_triggers) \
+	{ string, sizeof(string) - 1, ptype, eatype, is_attachable, atype, has_triggers }
 
 /* Programs that can NOT be attached. */
-#define BPF_PROG_SEC(string, ptype) BPF_PROG_SEC_IMPL(string, ptype, 0, 0, 0)
+#define BPF_PROG_SEC(string, ptype) BPF_PROG_SEC_IMPL(string, ptype, 0, 0, 0, false)
 
 /* Programs that can be attached. */
 #define BPF_APROG_SEC(string, ptype, atype) \
-	BPF_PROG_SEC_IMPL(string, ptype, 0, 1, atype)
+	BPF_PROG_SEC_IMPL(string, ptype, 0, 1, atype, false)
 
 /* Programs that must specify expected attach type at load time. */
 #define BPF_EAPROG_SEC(string, ptype, eatype) \
-	BPF_PROG_SEC_IMPL(string, ptype, eatype, 1, eatype)
+	BPF_PROG_SEC_IMPL(string, ptype, eatype, 1, eatype, false)
+
+/* Programs that must specify expected attach type at load time and has triggers. */
+#define BPF_TEAPROG_SEC(string, ptype, eatype) \
+	BPF_PROG_SEC_IMPL(string, ptype, eatype, 1, eatype, true)
 
 /* Programs that can be attached but attach type can't be identified by section
  * name. Kept for backward compatibility.
@@ -3566,6 +3578,7 @@ static const struct {
 	enum bpf_attach_type expected_attach_type;
 	int is_attachable;
 	enum bpf_attach_type attach_type;
+	bool has_triggers;
 } section_names[] = {
 	BPF_PROG_SEC("socket",			BPF_PROG_TYPE_SOCKET_FILTER),
 	BPF_PROG_SEC("kprobe/",			BPF_PROG_TYPE_KPROBE),
@@ -3628,6 +3641,10 @@ static const struct {
 						BPF_CGROUP_GETSOCKOPT),
 	BPF_EAPROG_SEC("cgroup/setsockopt",	BPF_PROG_TYPE_CGROUP_SOCKOPT,
 						BPF_CGROUP_SETSOCKOPT),
+	BPF_EAPROG_SEC("landlock/fs_walk",	BPF_PROG_TYPE_LANDLOCK_HOOK,
+						BPF_LANDLOCK_FS_WALK),
+	BPF_TEAPROG_SEC("landlock/fs_pick",	BPF_PROG_TYPE_LANDLOCK_HOOK,
+						BPF_LANDLOCK_FS_PICK),
 };
 
 #undef BPF_PROG_SEC_IMPL
@@ -3665,7 +3682,8 @@ static char *libbpf_get_type_names(bool attach_type)
 }
 
 int libbpf_prog_type_by_name(const char *name, enum bpf_prog_type *prog_type,
-			     enum bpf_attach_type *expected_attach_type)
+			     enum bpf_attach_type *expected_attach_type,
+			     bool *has_triggers)
 {
 	char *type_names;
 	int i;
@@ -3678,6 +3696,7 @@ int libbpf_prog_type_by_name(const char *name, enum bpf_prog_type *prog_type,
 			continue;
 		*prog_type = section_names[i].prog_type;
 		*expected_attach_type = section_names[i].expected_attach_type;
+		*has_triggers = section_names[i].has_triggers;
 		return 0;
 	}
 	pr_warning("failed to guess program type based on ELF section name '%s'\n", name);
@@ -3720,10 +3739,11 @@ int libbpf_attach_type_by_name(const char *name,
 static int
 bpf_program__identify_section(struct bpf_program *prog,
 			      enum bpf_prog_type *prog_type,
-			      enum bpf_attach_type *expected_attach_type)
+			      enum bpf_attach_type *expected_attach_type,
+			      bool *has_triggers)
 {
 	return libbpf_prog_type_by_name(prog->section_name, prog_type,
-					expected_attach_type);
+					expected_attach_type, has_triggers);
 }
 
 int bpf_map__fd(const struct bpf_map *map)
@@ -3898,6 +3918,7 @@ int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr,
 	struct bpf_object *obj;
 	struct bpf_map *map;
 	int err;
+	bool has_triggers = false;
 
 	if (!attr)
 		return -EINVAL;
@@ -3921,7 +3942,8 @@ int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr,
 		expected_attach_type = attr->expected_attach_type;
 		if (prog_type == BPF_PROG_TYPE_UNSPEC) {
 			err = bpf_program__identify_section(prog, &prog_type,
-							    &expected_attach_type);
+							    &expected_attach_type,
+							    &has_triggers);
 			if (err < 0) {
 				bpf_object__close(obj);
 				return -EINVAL;
@@ -3931,6 +3953,9 @@ int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr,
 		bpf_program__set_type(prog, prog_type);
 		bpf_program__set_expected_attach_type(prog,
 						      expected_attach_type);
+		if (has_triggers)
+			bpf_program__set_expected_attach_triggers(prog,
+					attr->expected_attach_triggers);
 
 		prog->log_level = attr->log_level;
 		prog->prog_flags = attr->prog_flags;
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 5cbf459ece0b..07e153cebd5d 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -123,7 +123,8 @@ LIBBPF_API void *bpf_object__priv(const struct bpf_object *prog);
 
 LIBBPF_API int
 libbpf_prog_type_by_name(const char *name, enum bpf_prog_type *prog_type,
-			 enum bpf_attach_type *expected_attach_type);
+			 enum bpf_attach_type *expected_attach_type,
+			 bool *has_triggers);
 LIBBPF_API int libbpf_attach_type_by_name(const char *name,
 					  enum bpf_attach_type *attach_type);
 
@@ -266,6 +267,9 @@ LIBBPF_API void bpf_program__set_type(struct bpf_program *prog,
 LIBBPF_API void
 bpf_program__set_expected_attach_type(struct bpf_program *prog,
 				      enum bpf_attach_type type);
+LIBBPF_API void
+bpf_program__set_expected_attach_triggers(struct bpf_program *prog,
+					  __u64 triggers);
 
 LIBBPF_API bool bpf_program__is_socket_filter(const struct bpf_program *prog);
 LIBBPF_API bool bpf_program__is_tracepoint(const struct bpf_program *prog);
@@ -345,6 +349,7 @@ struct bpf_prog_load_attr {
 	const char *file;
 	enum bpf_prog_type prog_type;
 	enum bpf_attach_type expected_attach_type;
+	__u64 expected_attach_triggers;
 	int ifindex;
 	int log_level;
 	int prog_flags;
diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
index 36ac26bdfda0..4eb930bfc1d8 100644
--- a/tools/lib/bpf/libbpf.map
+++ b/tools/lib/bpf/libbpf.map
@@ -83,6 +83,7 @@ LIBBPF_0.0.1 {
 		bpf_program__prev;
 		bpf_program__priv;
 		bpf_program__set_expected_attach_type;
+		bpf_program__set_expected_attach_triggers;
 		bpf_program__set_ifindex;
 		bpf_program__set_kprobe;
 		bpf_program__set_perf_event;
diff --git a/tools/testing/selftests/bpf/bpf_helpers.h b/tools/testing/selftests/bpf/bpf_helpers.h
index 5a3d92c8bec8..db2a84a88f5c 100644
--- a/tools/testing/selftests/bpf/bpf_helpers.h
+++ b/tools/testing/selftests/bpf/bpf_helpers.h
@@ -228,6 +228,8 @@ static void *(*bpf_sk_storage_get)(void *map, struct bpf_sock *sk,
 static int (*bpf_sk_storage_delete)(void *map, struct bpf_sock *sk) =
 	(void *)BPF_FUNC_sk_storage_delete;
 static int (*bpf_send_signal)(unsigned sig) = (void *)BPF_FUNC_send_signal;
+static void *(*bpf_inode_map_lookup_elem)(void *map, const void *key) =
+	(void *) BPF_FUNC_inode_map_lookup_elem;
 
 /* llvm builtin functions that eBPF C program may use to
  * emit BPF_LD_ABS and BPF_LD_IND instructions
diff --git a/tools/testing/selftests/bpf/test_section_names.c b/tools/testing/selftests/bpf/test_section_names.c
index 29833aeaf0de..2d08df9156bd 100644
--- a/tools/testing/selftests/bpf/test_section_names.c
+++ b/tools/testing/selftests/bpf/test_section_names.c
@@ -153,7 +153,7 @@ static int test_prog_type_by_name(const struct sec_name_test *test)
 	int rc;
 
 	rc = libbpf_prog_type_by_name(test->sec_name, &prog_type,
-				      &expected_attach_type);
+				      &expected_attach_type, false);
 
 	if (rc != test->expected_load.rc) {
 		warnx("prog: unexpected rc=%d for %s", rc, test->sec_name);
diff --git a/tools/testing/selftests/bpf/test_sockopt_multi.c b/tools/testing/selftests/bpf/test_sockopt_multi.c
index 4be3441db867..e499c91f2953 100644
--- a/tools/testing/selftests/bpf/test_sockopt_multi.c
+++ b/tools/testing/selftests/bpf/test_sockopt_multi.c
@@ -23,7 +23,7 @@ static int prog_attach(struct bpf_object *obj, int cgroup_fd, const char *title)
 	struct bpf_program *prog;
 	int err;
 
-	err = libbpf_prog_type_by_name(title, &prog_type, &attach_type);
+	err = libbpf_prog_type_by_name(title, &prog_type, &attach_type, false);
 	if (err) {
 		log_err("Failed to deduct types for %s BPF program", title);
 		return -1;
@@ -52,7 +52,7 @@ static int prog_detach(struct bpf_object *obj, int cgroup_fd, const char *title)
 	struct bpf_program *prog;
 	int err;
 
-	err = libbpf_prog_type_by_name(title, &prog_type, &attach_type);
+	err = libbpf_prog_type_by_name(title, &prog_type, &attach_type, false);
 	if (err)
 		return -1;
 
diff --git a/tools/testing/selftests/bpf/test_sockopt_sk.c b/tools/testing/selftests/bpf/test_sockopt_sk.c
index 036b652e5ca9..2d1ff616b139 100644
--- a/tools/testing/selftests/bpf/test_sockopt_sk.c
+++ b/tools/testing/selftests/bpf/test_sockopt_sk.c
@@ -129,7 +129,7 @@ static int prog_attach(struct bpf_object *obj, int cgroup_fd, const char *title)
 	struct bpf_program *prog;
 	int err;
 
-	err = libbpf_prog_type_by_name(title, &prog_type, &attach_type);
+	err = libbpf_prog_type_by_name(title, &prog_type, &attach_type, false);
 	if (err) {
 		log_err("Failed to deduct types for %s BPF program", title);
 		return -1;
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 09/10] bpf,landlock: Add tests for Landlock
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
                   ` (7 preceding siblings ...)
  2019-07-21 21:31 ` [PATCH bpf-next v10 08/10] bpf: Add a Landlock sandbox example Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-21 21:31 ` [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation " Mickaël Salaün
  9 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

Test basic context access, ptrace protection and filesystem hooks.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: James Morris <jmorris@namei.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Will Drewry <wad@chromium.org>
---

Changes since v9:
* replace subtype with expected_attach_type and expected_attach_triggers
* rename inode_map_lookup() into inode_map_lookup_elem()
* check for inode map entry without value (which is now possible thanks
  to the pointer null check)
* use read-only inode map for Landlock programs

Changes since v8:
* update eBPF include path for macros
* use TEST_GEN_PROGS and use the generic "clean" target
* add more verbose errors
* update the bpf/verifier files
* remove chain tests (from landlock and bpf/verifier)
* replace the whitelist tests with blacklist tests (because of stateless
  Landlock programs): remove "dotdot" tests and other depth tests
* sync the landlock Makefile with its bpf sibling directory and use
  bpf_load_program_xattr()

Changes since v7:
* update tests and add new ones for filesystem hierarchy and Landlock
  chains.

Changes since v6:
* use the new kselftest_harness.h
* use const variables
* replace ASSERT_STEP with ASSERT_*
* rename BPF_PROG_TYPE_LANDLOCK to BPF_PROG_TYPE_LANDLOCK_RULE
* force sample library rebuild
* fix install target

Changes since v5:
* add subtype test
* add ptrace tests
* split and rename files
* cleanup and rebase
---
 tools/testing/selftests/Makefile              |   1 +
 tools/testing/selftests/bpf/test_verifier.c   |   1 +
 .../testing/selftests/bpf/verifier/landlock.c |  24 ++
 tools/testing/selftests/landlock/.gitignore   |   4 +
 tools/testing/selftests/landlock/Makefile     |  39 +++
 tools/testing/selftests/landlock/test.h       |  50 ++++
 tools/testing/selftests/landlock/test_base.c  |  24 ++
 tools/testing/selftests/landlock/test_fs.c    | 256 ++++++++++++++++++
 .../testing/selftests/landlock/test_ptrace.c  | 148 ++++++++++
 9 files changed, 547 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/verifier/landlock.c
 create mode 100644 tools/testing/selftests/landlock/.gitignore
 create mode 100644 tools/testing/selftests/landlock/Makefile
 create mode 100644 tools/testing/selftests/landlock/test.h
 create mode 100644 tools/testing/selftests/landlock/test_base.c
 create mode 100644 tools/testing/selftests/landlock/test_fs.c
 create mode 100644 tools/testing/selftests/landlock/test_ptrace.c

diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile
index 25b43a8c2b15..1949fbb3098e 100644
--- a/tools/testing/selftests/Makefile
+++ b/tools/testing/selftests/Makefile
@@ -21,6 +21,7 @@ TARGETS += ir
 TARGETS += kcmp
 TARGETS += kexec
 TARGETS += kvm
+TARGETS += landlock
 TARGETS += lib
 TARGETS += livepatch
 TARGETS += membarrier
diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c
index b0773291012a..b8542431c78b 100644
--- a/tools/testing/selftests/bpf/test_verifier.c
+++ b/tools/testing/selftests/bpf/test_verifier.c
@@ -30,6 +30,7 @@
 #include <linux/bpf.h>
 #include <linux/if_ether.h>
 #include <linux/btf.h>
+#include <linux/landlock.h>
 
 #include <bpf/bpf.h>
 #include <bpf/libbpf.h>
diff --git a/tools/testing/selftests/bpf/verifier/landlock.c b/tools/testing/selftests/bpf/verifier/landlock.c
new file mode 100644
index 000000000000..eaf6dddbf208
--- /dev/null
+++ b/tools/testing/selftests/bpf/verifier/landlock.c
@@ -0,0 +1,24 @@
+{
+	"landlock/fs_walk: always accept",
+	.insns = {
+		BPF_MOV32_IMM(BPF_REG_0, 0),
+		BPF_EXIT_INSN(),
+	},
+	.result = ACCEPT,
+	.prog_type = BPF_PROG_TYPE_LANDLOCK_HOOK,
+	.expected_attach_type = BPF_LANDLOCK_FS_WALK,
+},
+{
+	"landlock/fs_pick: read context",
+	.insns = {
+		BPF_MOV64_REG(BPF_REG_6, BPF_REG_1),
+		BPF_LDX_MEM(BPF_DW, BPF_REG_7, BPF_REG_6,
+			offsetof(struct landlock_ctx_fs_pick, inode)),
+		BPF_MOV32_IMM(BPF_REG_0, 0),
+		BPF_EXIT_INSN(),
+	},
+	.result = ACCEPT,
+	.prog_type = BPF_PROG_TYPE_LANDLOCK_HOOK,
+	.expected_attach_type = BPF_LANDLOCK_FS_PICK,
+	.expected_attach_triggers = LANDLOCK_TRIGGER_FS_PICK_READ,
+},
diff --git a/tools/testing/selftests/landlock/.gitignore b/tools/testing/selftests/landlock/.gitignore
new file mode 100644
index 000000000000..25b9cd834c3c
--- /dev/null
+++ b/tools/testing/selftests/landlock/.gitignore
@@ -0,0 +1,4 @@
+/test_base
+/test_fs
+/test_ptrace
+/tmp_*
diff --git a/tools/testing/selftests/landlock/Makefile b/tools/testing/selftests/landlock/Makefile
new file mode 100644
index 000000000000..7a253bf6d580
--- /dev/null
+++ b/tools/testing/selftests/landlock/Makefile
@@ -0,0 +1,39 @@
+LIBDIR := ../../../lib
+BPFDIR := $(LIBDIR)/bpf
+APIDIR := ../../../include/uapi
+GENDIR := ../../../../include/generated
+GENHDR := $(GENDIR)/autoconf.h
+
+ifneq ($(wildcard $(GENHDR)),)
+  GENFLAGS := -DHAVE_GENHDR
+endif
+
+BPFOBJS := $(BPFDIR)/bpf.o $(BPFDIR)/nlattr.o
+LOADOBJ := ../../../../samples/bpf/bpf_load.o
+
+CFLAGS += -Wl,-no-as-needed -Wall -O2 -I$(APIDIR) -I$(LIBDIR) -I$(BPFDIR) -I$(GENDIR) $(GENFLAGS) -I../../../include
+LDFLAGS += -lelf
+
+test_src = $(wildcard test_*.c)
+
+test_objs := $(test_src:.c=)
+
+TEST_GEN_PROGS := $(test_objs)
+
+.PHONY: all clean force
+
+all: $(test_objs)
+
+# force a rebuild of BPFOBJS when its dependencies are updated
+force:
+
+# rebuild bpf.o as a workaround for the samples/bpf bug
+$(BPFOBJS): $(LOADOBJ) force
+	$(MAKE) -C $(BPFDIR)
+
+$(LOADOBJ): force
+	$(MAKE) -C $(dir $(LOADOBJ))
+
+$(test_objs): $(BPFOBJS) $(LOADOBJ) ../kselftest_harness.h
+
+include ../lib.mk
diff --git a/tools/testing/selftests/landlock/test.h b/tools/testing/selftests/landlock/test.h
new file mode 100644
index 000000000000..e1e86a804180
--- /dev/null
+++ b/tools/testing/selftests/landlock/test.h
@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Landlock helpers
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ * Copyright © 2019 ANSSI
+ */
+
+#include <bpf/bpf.h>
+#include <errno.h>
+#include <linux/filter.h>
+#include <linux/landlock.h>
+#include <linux/seccomp.h>
+#include <sys/prctl.h>
+#include <sys/syscall.h>
+
+#include "../kselftest_harness.h"
+#include "../../../../samples/bpf/bpf_load.h"
+
+#ifndef SECCOMP_PREPEND_LANDLOCK_PROG
+#define SECCOMP_PREPEND_LANDLOCK_PROG	4
+#endif
+
+#ifndef seccomp
+static int __attribute__((unused)) seccomp(unsigned int op, unsigned int flags,
+		void *args)
+{
+	errno = 0;
+	return syscall(__NR_seccomp, op, flags, args);
+}
+#endif
+
+/* bpf_load_program() with subtype */
+static int __attribute__((unused)) ll_bpf_load_program(
+		const struct bpf_insn *insns, size_t insns_cnt, char *log_buf,
+		size_t log_buf_sz, const enum bpf_attach_type attach_type,
+		__u64 attach_triggers)
+{
+	struct bpf_load_program_attr load_attr;
+
+	memset(&load_attr, 0, sizeof(struct bpf_load_program_attr));
+	load_attr.prog_type = BPF_PROG_TYPE_LANDLOCK_HOOK;
+	load_attr.expected_attach_type = attach_type;
+	load_attr.expected_attach_triggers = attach_triggers;
+	load_attr.insns = insns;
+	load_attr.insns_cnt = insns_cnt;
+	load_attr.license = "GPL";
+
+	return bpf_load_program_xattr(&load_attr, log_buf, log_buf_sz);
+}
diff --git a/tools/testing/selftests/landlock/test_base.c b/tools/testing/selftests/landlock/test_base.c
new file mode 100644
index 000000000000..db46f39048cb
--- /dev/null
+++ b/tools/testing/selftests/landlock/test_base.c
@@ -0,0 +1,24 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Landlock tests - base
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ */
+
+#define _GNU_SOURCE
+#include <errno.h>
+
+#include "test.h"
+
+TEST(seccomp_landlock)
+{
+	int ret;
+
+	ret = seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, NULL);
+	EXPECT_EQ(-1, ret);
+	EXPECT_EQ(EFAULT, errno) {
+		TH_LOG("Kernel does not support CONFIG_SECURITY_LANDLOCK");
+	}
+}
+
+TEST_HARNESS_MAIN
diff --git a/tools/testing/selftests/landlock/test_fs.c b/tools/testing/selftests/landlock/test_fs.c
new file mode 100644
index 000000000000..f35b99fcb70f
--- /dev/null
+++ b/tools/testing/selftests/landlock/test_fs.c
@@ -0,0 +1,256 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Landlock tests - file system
+ *
+ * Copyright © 2018-2019 Mickaël Salaün <mic@digikod.net>
+ */
+
+#include <bpf/bpf.h> /* bpf_create_map() */
+#include <fcntl.h> /* O_DIRECTORY */
+#include <sys/stat.h> /* statbuf */
+#include <unistd.h> /* faccessat() */
+
+#include "test.h"
+
+#define TEST_PATH_TRIGGERS ( \
+		LANDLOCK_TRIGGER_FS_PICK_OPEN | \
+		LANDLOCK_TRIGGER_FS_PICK_READDIR | \
+		LANDLOCK_TRIGGER_FS_PICK_EXECUTE | \
+		LANDLOCK_TRIGGER_FS_PICK_GETATTR)
+
+static void test_path_rel(struct __test_metadata *_metadata, int dirfd,
+		const char *path, int ret)
+{
+	int fd;
+	struct stat statbuf;
+
+	ASSERT_EQ(ret, faccessat(dirfd, path, R_OK | X_OK, 0));
+	ASSERT_EQ(ret, fstatat(dirfd, path, &statbuf, 0));
+	fd = openat(dirfd, path, O_DIRECTORY);
+	if (ret) {
+		ASSERT_EQ(-1, fd);
+	} else {
+		ASSERT_NE(-1, fd);
+		EXPECT_EQ(0, close(fd));
+	}
+}
+
+static void test_path(struct __test_metadata *_metadata, const char *path,
+		int ret)
+{
+	return test_path_rel(_metadata, AT_FDCWD, path, ret);
+}
+
+static const char d1[] = "/usr";
+static const char d2[] = "/usr/share";
+static const char d3[] = "/usr/share/doc";
+
+TEST(fs_base)
+{
+	test_path(_metadata, d1, 0);
+	test_path(_metadata, d2, 0);
+	test_path(_metadata, d3, 0);
+}
+
+#define MAP_VALUE_DENY 1
+
+static int create_denied_inode_map(struct __test_metadata *_metadata,
+		const char *const dirs[])
+{
+	int map, key, dirs_len, i;
+	__u64 value = MAP_VALUE_DENY;
+
+	ASSERT_NE(NULL, dirs) {
+		TH_LOG("No directory list\n");
+	}
+	ASSERT_NE(NULL, dirs[0]) {
+		TH_LOG("Empty directory list\n");
+	}
+
+	/* get the number of dir entries */
+	for (dirs_len = 0; dirs[dirs_len]; dirs_len++);
+	map = bpf_create_map(BPF_MAP_TYPE_INODE, sizeof(key), sizeof(value),
+			dirs_len, BPF_F_RDONLY_PROG);
+	ASSERT_NE(-1, map) {
+		TH_LOG("Failed to create a map of %d elements: %s\n", dirs_len,
+				strerror(errno));
+	}
+
+	for (i = 0; dirs[i]; i++) {
+		key = open(dirs[i], O_RDONLY | O_CLOEXEC | O_DIRECTORY);
+		ASSERT_NE(-1, key) {
+			TH_LOG("Failed to open directory \"%s\": %s\n", dirs[i],
+					strerror(errno));
+		}
+		ASSERT_EQ(0, bpf_map_update_elem(map, &key, &value, BPF_ANY)) {
+			TH_LOG("Failed to update the map with \"%s\": %s\n",
+					dirs[i], strerror(errno));
+		}
+		close(key);
+	}
+	return map;
+}
+
+static void enforce_map(struct __test_metadata *_metadata, int map,
+		bool subpath)
+{
+	const struct bpf_insn prog_deny[] = {
+		BPF_ALU64_REG(BPF_MOV, BPF_REG_6, BPF_REG_1),
+		/* look for the requested inode in the map */
+		BPF_LDX_MEM(BPF_DW, BPF_REG_2, BPF_REG_6,
+			offsetof(struct landlock_ctx_fs_walk, inode)),
+		BPF_LD_MAP_FD(BPF_REG_1, map), /* 2 instructions */
+		BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
+				BPF_FUNC_inode_map_lookup_elem),
+		/* if there is no mark, then allow access to this inode */
+		BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 2),
+		BPF_LDX_MEM(BPF_DW, BPF_REG_1, BPF_REG_0, 0),
+		/* otherwise, deny access to this inode */
+		BPF_JMP_IMM(BPF_JEQ, BPF_REG_1, MAP_VALUE_DENY, 2),
+		BPF_MOV32_IMM(BPF_REG_0, LANDLOCK_RET_ALLOW),
+		BPF_EXIT_INSN(),
+		BPF_MOV32_IMM(BPF_REG_0, LANDLOCK_RET_DENY),
+		BPF_EXIT_INSN(),
+	};
+	int fd_walk = -1, fd_pick;
+	char log[1024] = "";
+
+	if (subpath) {
+		fd_walk = ll_bpf_load_program((const struct bpf_insn *)&prog_deny,
+				sizeof(prog_deny) / sizeof(struct bpf_insn),
+				log, sizeof(log), BPF_LANDLOCK_FS_WALK, 0);
+		ASSERT_NE(-1, fd_walk) {
+			TH_LOG("Failed to load fs_walk program: %s\n%s",
+					strerror(errno), log);
+		}
+		ASSERT_EQ(0, seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, &fd_walk)) {
+			TH_LOG("Failed to apply Landlock program: %s", strerror(errno));
+		}
+		EXPECT_EQ(0, close(fd_walk));
+	}
+
+	fd_pick = ll_bpf_load_program((const struct bpf_insn *)&prog_deny,
+			sizeof(prog_deny) / sizeof(struct bpf_insn), log,
+			sizeof(log), BPF_LANDLOCK_FS_PICK, TEST_PATH_TRIGGERS);
+	ASSERT_NE(-1, fd_pick) {
+		TH_LOG("Failed to load fs_pick program: %s\n%s",
+				strerror(errno), log);
+	}
+	ASSERT_EQ(0, seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, &fd_pick)) {
+		TH_LOG("Failed to apply Landlock program: %s", strerror(errno));
+	}
+	EXPECT_EQ(0, close(fd_pick));
+}
+
+static void check_map_blacklist(struct __test_metadata *_metadata,
+		bool subpath)
+{
+	int map = create_denied_inode_map(_metadata, (const char *const [])
+			{ d2, NULL });
+	ASSERT_NE(-1, map);
+	enforce_map(_metadata, map, subpath);
+	test_path(_metadata, d1, 0);
+	test_path(_metadata, d2, -1);
+	test_path(_metadata, d3, subpath ? -1 : 0);
+	EXPECT_EQ(0, close(map));
+}
+
+TEST(fs_map_blacklist_literal)
+{
+	check_map_blacklist(_metadata, false);
+}
+
+TEST(fs_map_blacklist_subpath)
+{
+	check_map_blacklist(_metadata, true);
+}
+
+static const char r2[] = ".";
+static const char r3[] = "./doc";
+
+enum relative_access {
+	REL_OPEN,
+	REL_CHDIR,
+	REL_CHROOT,
+};
+
+static void check_access(struct __test_metadata *_metadata,
+		bool enforce, enum relative_access rel)
+{
+	int dirfd;
+	int map = -1;
+
+	if (rel == REL_CHROOT)
+		ASSERT_NE(-1, chdir(d2));
+	if (enforce) {
+		map = create_denied_inode_map(_metadata, (const char *const [])
+				{ d3, NULL });
+		ASSERT_NE(-1, map);
+		enforce_map(_metadata, map, true);
+	}
+	switch (rel) {
+	case REL_OPEN:
+		dirfd = open(d2, O_DIRECTORY);
+		ASSERT_NE(-1, dirfd);
+		break;
+	case REL_CHDIR:
+		ASSERT_NE(-1, chdir(d2));
+		dirfd = AT_FDCWD;
+		break;
+	case REL_CHROOT:
+		ASSERT_NE(-1, chroot(d2)) {
+			TH_LOG("Failed to chroot: %s\n", strerror(errno));
+		}
+		dirfd = AT_FDCWD;
+		break;
+	default:
+		ASSERT_TRUE(false);
+		return;
+	}
+
+	test_path_rel(_metadata, dirfd, r2, 0);
+	test_path_rel(_metadata, dirfd, r3, enforce ? -1 : 0);
+
+	if (rel == REL_OPEN)
+		EXPECT_EQ(0, close(dirfd));
+	if (enforce)
+		EXPECT_EQ(0, close(map));
+}
+
+TEST(fs_allow_open)
+{
+	/* no enforcement, via open */
+	check_access(_metadata, false, REL_OPEN);
+}
+
+TEST(fs_allow_chdir)
+{
+	/* no enforcement, via chdir */
+	check_access(_metadata, false, REL_CHDIR);
+}
+
+TEST(fs_allow_chroot)
+{
+	/* no enforcement, via chroot */
+	check_access(_metadata, false, REL_CHROOT);
+}
+
+TEST(fs_deny_open)
+{
+	/* enforcement without tag, via open */
+	check_access(_metadata, true, REL_OPEN);
+}
+
+TEST(fs_deny_chdir)
+{
+	/* enforcement without tag, via chdir */
+	check_access(_metadata, true, REL_CHDIR);
+}
+
+TEST(fs_deny_chroot)
+{
+	/* enforcement without tag, via chroot */
+	check_access(_metadata, true, REL_CHROOT);
+}
+
+TEST_HARNESS_MAIN
diff --git a/tools/testing/selftests/landlock/test_ptrace.c b/tools/testing/selftests/landlock/test_ptrace.c
new file mode 100644
index 000000000000..b190a809ceec
--- /dev/null
+++ b/tools/testing/selftests/landlock/test_ptrace.c
@@ -0,0 +1,148 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Landlock tests - ptrace
+ *
+ * Copyright © 2017-2019 Mickaël Salaün <mic@digikod.net>
+ */
+
+#define _GNU_SOURCE
+#include <signal.h> /* raise */
+#include <sys/ptrace.h>
+#include <sys/types.h> /* waitpid */
+#include <sys/wait.h> /* waitpid */
+#include <unistd.h> /* fork, pipe */
+
+#include "test.h"
+
+static void apply_null_sandbox(struct __test_metadata *_metadata)
+{
+	const struct bpf_insn prog_accept[] = {
+		BPF_MOV32_IMM(BPF_REG_0, 0),
+		BPF_EXIT_INSN(),
+	};
+	int prog;
+	char log[256] = "";
+
+	prog = ll_bpf_load_program((const struct bpf_insn *)&prog_accept,
+			sizeof(prog_accept) / sizeof(struct bpf_insn), log,
+			sizeof(log), BPF_LANDLOCK_FS_PICK, LANDLOCK_TRIGGER_FS_PICK_OPEN);
+	ASSERT_NE(-1, prog) {
+		TH_LOG("Failed to load minimal rule: %s\n%s",
+				strerror(errno), log);
+	}
+	ASSERT_EQ(0, seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, &prog)) {
+		TH_LOG("Failed to apply minimal rule: %s", strerror(errno));
+	}
+	EXPECT_EQ(0, close(prog));
+}
+
+/* PTRACE_TRACEME and PTRACE_ATTACH without Landlock rules effect */
+static void check_ptrace(struct __test_metadata *_metadata,
+		int sandbox_both, int sandbox_parent, int sandbox_child,
+		int expect_ptrace)
+{
+	pid_t child;
+	int status;
+	int pipefd[2];
+
+	ASSERT_EQ(0, pipe(pipefd));
+	if (sandbox_both)
+		apply_null_sandbox(_metadata);
+
+	child = fork();
+	ASSERT_LE(0, child);
+	if (child == 0) {
+		char buf;
+
+		EXPECT_EQ(0, close(pipefd[1]));
+		if (sandbox_child)
+			apply_null_sandbox(_metadata);
+
+		/* test traceme */
+		ASSERT_EQ(expect_ptrace, ptrace(PTRACE_TRACEME));
+		if (expect_ptrace) {
+			ASSERT_EQ(EPERM, errno);
+		} else {
+			ASSERT_EQ(0, raise(SIGSTOP));
+		}
+
+		/* sync */
+		ASSERT_EQ(1, read(pipefd[0], &buf, 1)) {
+			TH_LOG("Failed to read() sync from parent");
+		}
+		ASSERT_EQ('.', buf);
+		_exit(_metadata->passed ? EXIT_SUCCESS : EXIT_FAILURE);
+	}
+
+	EXPECT_EQ(0, close(pipefd[0]));
+	if (sandbox_parent)
+		apply_null_sandbox(_metadata);
+
+	/* test traceme */
+	if (!expect_ptrace) {
+		ASSERT_EQ(child, waitpid(child, &status, 0));
+		ASSERT_EQ(1, WIFSTOPPED(status));
+		ASSERT_EQ(0, ptrace(PTRACE_DETACH, child, NULL, 0));
+	}
+	/* test attach */
+	ASSERT_EQ(expect_ptrace, ptrace(PTRACE_ATTACH, child, NULL, 0));
+	if (expect_ptrace) {
+		ASSERT_EQ(EPERM, errno);
+	} else {
+		ASSERT_EQ(child, waitpid(child, &status, 0));
+		ASSERT_EQ(1, WIFSTOPPED(status));
+		ASSERT_EQ(0, ptrace(PTRACE_CONT, child, NULL, 0));
+	}
+
+	/* sync */
+	ASSERT_EQ(1, write(pipefd[1], ".", 1)) {
+		TH_LOG("Failed to write() sync to child");
+	}
+	ASSERT_EQ(child, waitpid(child, &status, 0));
+	if (WIFSIGNALED(status) || WEXITSTATUS(status))
+		_metadata->passed = 0;
+}
+
+TEST(ptrace_allow_without_sandbox)
+{
+	/* no sandbox */
+	check_ptrace(_metadata, 0, 0, 0, 0);
+}
+
+TEST(ptrace_allow_with_one_sandbox)
+{
+	/* child sandbox */
+	check_ptrace(_metadata, 0, 0, 1, 0);
+}
+
+TEST(ptrace_allow_with_nested_sandbox)
+{
+	/* inherited and child sandbox */
+	check_ptrace(_metadata, 1, 0, 1, 0);
+}
+
+TEST(ptrace_deny_with_parent_sandbox)
+{
+	/* parent sandbox */
+	check_ptrace(_metadata, 0, 1, 0, -1);
+}
+
+TEST(ptrace_deny_with_nested_and_parent_sandbox)
+{
+	/* inherited and parent sandbox */
+	check_ptrace(_metadata, 1, 1, 0, -1);
+}
+
+TEST(ptrace_deny_with_forked_sandbox)
+{
+	/* inherited, parent and child sandbox */
+	check_ptrace(_metadata, 1, 1, 1, -1);
+}
+
+TEST(ptrace_deny_with_sibling_sandbox)
+{
+	/* parent and child sandbox */
+	check_ptrace(_metadata, 0, 1, 1, -1);
+}
+
+TEST_HARNESS_MAIN
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation for Landlock
  2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
                   ` (8 preceding siblings ...)
  2019-07-21 21:31 ` [PATCH bpf-next v10 09/10] bpf,landlock: Add tests for Landlock Mickaël Salaün
@ 2019-07-21 21:31 ` Mickaël Salaün
  2019-07-31  1:53   ` Randy Dunlap
  9 siblings, 1 reply; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-21 21:31 UTC (permalink / raw)
  To: linux-kernel
  Cc: Mickaël Salaün, Alexander Viro, Alexei Starovoitov,
	Andrew Morton, Andy Lutomirski, Arnaldo Carvalho de Melo,
	Casey Schaufler, Daniel Borkmann, David Drysdale,
	David S . Miller, Eric W . Biederman, James Morris, Jann Horn,
	John Johansen, Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

This documentation can be built with the Sphinx framework.

Signed-off-by: Mickaël Salaün <mic@digikod.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: James Morris <jmorris@namei.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge E. Hallyn <serge@hallyn.com>
---

Changes since v9:
* update with expected attach type and expected attach triggers

Changes since v8:
* remove documentation related to chaining and tagging according to this
  patch series

Changes since v7:
* update documentation according to the Landlock revamp

Changes since v6:
* add a check for ctx->event
* rename BPF_PROG_TYPE_LANDLOCK to BPF_PROG_TYPE_LANDLOCK_RULE
* rename Landlock version to ABI to better reflect its purpose and add a
  dedicated changelog section
* update tables
* relax no_new_privs recommendations
* remove ABILITY_WRITE related functions
* reword rule "appending" to "prepending" and explain it
* cosmetic fixes

Changes since v5:
* update the rule hierarchy inheritance explanation
* briefly explain ctx->arg2
* add ptrace restrictions
* explain EPERM
* update example (subtype)
* use ":manpage:"
---
 Documentation/security/index.rst           |   1 +
 Documentation/security/landlock/index.rst  |  20 +++
 Documentation/security/landlock/kernel.rst |  99 ++++++++++++++
 Documentation/security/landlock/user.rst   | 147 +++++++++++++++++++++
 4 files changed, 267 insertions(+)
 create mode 100644 Documentation/security/landlock/index.rst
 create mode 100644 Documentation/security/landlock/kernel.rst
 create mode 100644 Documentation/security/landlock/user.rst

diff --git a/Documentation/security/index.rst b/Documentation/security/index.rst
index aad6d92ffe31..32b4c1db2325 100644
--- a/Documentation/security/index.rst
+++ b/Documentation/security/index.rst
@@ -12,3 +12,4 @@ Security Documentation
    SCTP
    self-protection
    tpm/index
+   landlock/index
diff --git a/Documentation/security/landlock/index.rst b/Documentation/security/landlock/index.rst
new file mode 100644
index 000000000000..d0af868d1582
--- /dev/null
+++ b/Documentation/security/landlock/index.rst
@@ -0,0 +1,20 @@
+=========================================
+Landlock LSM: programmatic access control
+=========================================
+
+Landlock is a stackable Linux Security Module (LSM) that makes it possible to
+create security sandboxes, programmable access-controls or safe endpoint
+security agents.  This kind of sandbox is expected to help mitigate the
+security impact of bugs or unexpected/malicious behaviors in user-space
+applications.  The current version allows only a process with the global
+CAP_SYS_ADMIN capability to create such sandboxes but the ultimate goal of
+Landlock is to empower any process, including unprivileged ones, to securely
+restrict themselves.  Landlock is inspired by seccomp-bpf but instead of
+filtering syscalls and their raw arguments, a Landlock rule can inspect the use
+of kernel objects like files and hence make a decision according to the kernel
+semantic.
+
+.. toctree::
+
+    user
+    kernel
diff --git a/Documentation/security/landlock/kernel.rst b/Documentation/security/landlock/kernel.rst
new file mode 100644
index 000000000000..7d1e06d544bf
--- /dev/null
+++ b/Documentation/security/landlock/kernel.rst
@@ -0,0 +1,99 @@
+==============================
+Landlock: kernel documentation
+==============================
+
+eBPF properties
+===============
+
+To get an expressive language while still being safe and small, Landlock is
+based on eBPF. Landlock should be usable by untrusted processes and must
+therefore expose a minimal attack surface. The eBPF bytecode is minimal,
+powerful, widely used and designed to be used by untrusted applications. Thus,
+reusing the eBPF support in the kernel enables a generic approach while
+minimizing new code.
+
+An eBPF program has access to an eBPF context containing some fields used to
+inspect the current object. These arguments can be used directly (e.g. cookie)
+or passed to helper functions according to their types (e.g. inode pointer). It
+is then possible to do complex access checks without race conditions or
+inconsistent evaluation (i.e.  `incorrect mirroring of the OS code and state
+<https://www.ndss-symposium.org/ndss2003/traps-and-pitfalls-practical-problems-system-call-interposition-based-security-tools/>`_).
+
+A Landlock hook describes a particular access type.  For now, there is two
+hooks dedicated to filesystem related operations: LANDLOCK_HOOK_FS_PICK and
+LANDLOCK_HOOK_FS_WALK.  A Landlock program is tied to one hook.  This makes it
+possible to statically check context accesses, potentially performed by such
+program, and hence prevents kernel address leaks and ensure the right use of
+hook arguments with eBPF functions.  Any user can add multiple Landlock
+programs per Landlock hook.  They are stacked and evaluated one after the
+other, starting from the most recent program, as seccomp-bpf does with its
+filters.  Underneath, a hook is an abstraction over a set of LSM hooks.
+
+
+Guiding principles
+==================
+
+Unprivileged use
+----------------
+
+* Landlock helpers and context should be usable by any unprivileged and
+  untrusted program while following the system security policy enforced by
+  other access control mechanisms (e.g. DAC, LSM).
+
+
+Landlock hook and context
+-------------------------
+
+* A Landlock hook shall be focused on access control on kernel objects instead
+  of syscall filtering (i.e. syscall arguments), which is the purpose of
+  seccomp-bpf.
+* A Landlock context provided by a hook shall express the minimal and more
+  generic interface to control an access for a kernel object.
+* A hook shall guaranty that all the BPF function calls from a program are
+  safe.  Thus, the related Landlock context arguments shall always be of the
+  same type for a particular hook.  For example, a network hook could share
+  helpers with a file hook because of UNIX socket.  However, the same helpers
+  may not be compatible for a file system handle and a net handle.
+* Multiple hooks may use the same context interface.
+
+
+Landlock helpers
+----------------
+
+* Landlock helpers shall be as generic as possible while at the same time being
+  as simple as possible and following the syscall creation principles (cf.
+  *Documentation/adding-syscalls.txt*).
+* The only behavior change allowed on a helper is to fix a (logical) bug to
+  match the initial semantic.
+* Helpers shall be reentrant, i.e. only take inputs from arguments (e.g. from
+  the BPF context), to enable a hook to use a cache.  Future program options
+  might change this cache behavior.
+* It is quite easy to add new helpers to extend Landlock.  The main concern
+  should be about the possibility to leak information from the kernel that may
+  not be accessible otherwise (i.e. side-channel attack).
+
+
+Questions and answers
+=====================
+
+Why not create a custom hook for each kind of action?
+-----------------------------------------------------
+
+Landlock programs can handle these checks.  Adding more exceptions to the
+kernel code would lead to more code complexity.  A decision to ignore a kind of
+action can and should be done at the beginning of a Landlock program.
+
+
+Why a program does not return an errno or a kill code?
+------------------------------------------------------
+
+seccomp filters can return multiple kind of code, including an errno value or a
+kill signal, which may be convenient for access control.  Those return codes
+are hardwired in the userland ABI.  Instead, Landlock's approach is to return a
+boolean to allow or deny an action, which is much simpler and more generic.
+Moreover, we do not really have a choice because, unlike to seccomp, Landlock
+programs are not enforced at the syscall entry point but may be executed at any
+point in the kernel (through LSM hooks) where an errno return code may not make
+sense.  However, with this simple ABI and with the ability to call helpers,
+Landlock may gain features similar to seccomp-bpf in the future while being
+compatible with previous programs.
diff --git a/Documentation/security/landlock/user.rst b/Documentation/security/landlock/user.rst
new file mode 100644
index 000000000000..14c4f3b377bd
--- /dev/null
+++ b/Documentation/security/landlock/user.rst
@@ -0,0 +1,147 @@
+================================
+Landlock: userland documentation
+================================
+
+Landlock programs
+=================
+
+eBPF programs are used to create security programs.  They are contained and can
+call only a whitelist of dedicated functions. Moreover, they can only loop
+under strict conditions, which protects from denial of service.  More
+information on BPF can be found in *Documentation/networking/filter.txt*.
+
+
+Writing a program
+-----------------
+
+To enforce a security policy, a thread first needs to create a Landlock program.
+The easiest way to write an eBPF program depicting a security program is to write
+it in the C language.  As described in *samples/bpf/README.rst*, LLVM can
+compile such programs.  Files *samples/bpf/landlock1_kern.c* and those in
+*tools/testing/selftests/landlock/* can be used as examples.
+
+Once the eBPF program is created, the next step is to create the metadata
+describing the Landlock program.  This metadata includes an expected attach type which
+contains the hook type to which the program is tied, and expected attach
+triggers which identify the actions for which the program should be run.
+
+A hook is a policy decision point which exposes the same context type for
+each program evaluation.
+
+A Landlock hook describes the kind of kernel object for which a program will be
+triggered to allow or deny an action.  For example, the hook
+BPF_LANDLOCK_FS_PICK can be triggered every time a landlocked thread performs a
+set of action related to the filesystem (e.g. open, read, write, mount...).
+This actions are identified by the `triggers` bitfield.
+
+The next step is to fill a :c:type:`struct bpf_load_program_attr
+<bpf_load_program_attr>` with BPF_PROG_TYPE_LANDLOCK_HOOK, the expected attach
+type and other BPF program metadata.  This bpf_attr must then be passed to the
+:manpage:`bpf(2)` syscall alongside the BPF_PROG_LOAD command.  If everything
+is deemed correct by the kernel, the thread gets a file descriptor referring to
+this program.
+
+In the following code, the *insn* variable is an array of BPF instructions
+which can be extracted from an ELF file as is done in bpf_load_file() from
+*samples/bpf/bpf_load.c*.
+
+.. code-block:: c
+
+    int prog_fd;
+    struct bpf_load_program_attr load_attr;
+
+    memset(&load_attr, 0, sizeof(struct bpf_load_program_attr));
+    load_attr.prog_type = BPF_PROG_TYPE_LANDLOCK_HOOK;
+    load_attr.expected_attach_type = BPF_LANDLOCK_FS_PICK;
+    load_attr.expected_attach_triggers = LANDLOCK_TRIGGER_FS_PICK_OPEN;
+    load_attr.insns = insns;
+    load_attr.insns_cnt = sizeof(insn) / sizeof(struct bpf_insn);
+    load_attr.license = "GPL";
+
+    prog_fd = bpf_load_program_xattr(&load_attr, log_buf, log_buf_sz);
+    if (prog_fd == -1)
+        exit(1);
+
+
+Enforcing a program
+-------------------
+
+Once the Landlock program has been created or received (e.g. through a UNIX
+socket), the thread willing to sandbox itself (and its future children) should
+perform the following two steps.
+
+The thread should first request to never be allowed to get new privileges with a
+call to :manpage:`prctl(2)` and the PR_SET_NO_NEW_PRIVS option.  More
+information can be found in *Documentation/prctl/no_new_privs.txt*.
+
+.. code-block:: c
+
+    if (prctl(PR_SET_NO_NEW_PRIVS, 1, NULL, 0, 0))
+        exit(1);
+
+A thread can apply a program to itself by using the :manpage:`seccomp(2)` syscall.
+The operation is SECCOMP_PREPEND_LANDLOCK_PROG, the flags must be empty and the
+*args* argument must point to a valid Landlock program file descriptor.
+
+.. code-block:: c
+
+    if (seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, &fd))
+        exit(1);
+
+If the syscall succeeds, the program is now enforced on the calling thread and
+will be enforced on all its subsequently created children of the thread as
+well.  Once a thread is landlocked, there is no way to remove this security
+policy, only stacking more restrictions is allowed.  The program evaluation is
+performed from the newest to the oldest.
+
+When a syscall ask for an action on a kernel object, if this action is denied,
+then an EACCES errno code is returned through the syscall.
+
+
+.. _inherited_programs:
+
+Inherited programs
+------------------
+
+Every new thread resulting from a :manpage:`clone(2)` inherits Landlock program
+restrictions from its parent.  This is similar to the seccomp inheritance as
+described in *Documentation/prctl/seccomp_filter.txt*.
+
+
+Ptrace restrictions
+-------------------
+
+A landlocked process has less privileges than a non-landlocked process and must
+then be subject to additional restrictions when manipulating another process.
+To be allowed to use :manpage:`ptrace(2)` and related syscalls on a target
+process, a landlocked process must have a subset of the target process programs.
+
+
+Landlock structures and constants
+=================================
+
+Hook types
+----------
+
+.. kernel-doc:: include/uapi/linux/landlock.h
+    :functions: landlock_hook_type
+
+
+Contexts
+--------
+
+.. kernel-doc:: include/uapi/linux/landlock.h
+    :functions: landlock_ctx_fs_pick landlock_ctx_fs_walk landlock_ctx_fs_get
+
+
+Triggers for fs_pick
+--------------------
+
+.. kernel-doc:: include/uapi/linux/landlock.h
+    :functions: landlock_triggers
+
+
+Additional documentation
+========================
+
+See https://landlock.io
-- 
2.22.0


^ permalink raw reply related	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-07-21 21:31 ` [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode Mickaël Salaün
@ 2019-07-27  1:40   ` Alexei Starovoitov
  2019-07-31 18:46     ` Mickaël Salaün
  0 siblings, 1 reply; 22+ messages in thread
From: Alexei Starovoitov @ 2019-07-27  1:40 UTC (permalink / raw)
  To: Mickaël Salaün
  Cc: linux-kernel, Alexander Viro, Alexei Starovoitov, Andrew Morton,
	Andy Lutomirski, Arnaldo Carvalho de Melo, Casey Schaufler,
	Daniel Borkmann, David Drysdale, David S . Miller,
	Eric W . Biederman, James Morris, Jann Horn, John Johansen,
	Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

On Sun, Jul 21, 2019 at 11:31:12PM +0200, Mickaël Salaün wrote:
> FIXME: 64-bits in the doc
> 
> This new map store arbitrary values referenced by inode keys.  The map
> can be updated from user space with file descriptor pointing to inodes
> tied to a file system.  From an eBPF (Landlock) program point of view,
> such a map is read-only and can only be used to retrieved a value tied
> to a given inode.  This is useful to recognize an inode tagged by user
> space, without access right to this inode (i.e. no need to have a write
> access to this inode).
> 
> Add dedicated BPF functions to handle this type of map:
> * bpf_inode_htab_map_update_elem()
> * bpf_inode_htab_map_lookup_elem()
> * bpf_inode_htab_map_delete_elem()
> 
> This new map require a dedicated helper inode_map_lookup_elem() because
> of the key which is a pointer to an opaque data (only provided by the
> kernel).  This act like a (physical or cryptographic) key, which is why
> it is also not allowed to get the next key.
> 
> Signed-off-by: Mickaël Salaün <mic@digikod.net>

there are too many things to comment on.
Let's do this patch.

imo inode_map concept is interesting, but see below...

> +
> +	/*
> +	 * Limit number of entries in an inode map to the maximum number of
> +	 * open files for the current process. The maximum number of file
> +	 * references (including all inode maps) for a process is then
> +	 * (RLIMIT_NOFILE - 1) * RLIMIT_NOFILE. If the process' RLIMIT_NOFILE
> +	 * is 0, then any entry update is forbidden.
> +	 *
> +	 * An eBPF program can inherit all the inode map FD. The worse case is
> +	 * to fill a bunch of arraymaps, create an eBPF program, close the
> +	 * inode map FDs, and start again. The maximum number of inode map
> +	 * entries can then be close to RLIMIT_NOFILE^3.
> +	 */
> +	if (attr->max_entries > rlimit(RLIMIT_NOFILE))
> +		return -EMFILE;

rlimit is checked, but no fd are consumed.
Once created such inode map_fd can be passed to a different process.
map_fd can be pinned into bpffs.
etc.
what the value of the check?

> +
> +	/* decorelate UAPI from kernel API */
> +	attr->key_size = sizeof(struct inode *);
> +
> +	return htab_map_alloc_check(attr);
> +}
> +
> +static void inode_htab_put_key(void *key)
> +{
> +	struct inode **inode = key;
> +
> +	if ((*inode)->i_state & I_FREEING)
> +		return;

checking the state without take a lock? isn't it racy?

> +	iput(*inode);
> +}
> +
> +/* called from syscall or (never) from eBPF program */
> +static int map_get_next_no_key(struct bpf_map *map, void *key, void *next_key)
> +{
> +	/* do not leak a file descriptor */

what this comment suppose to mean?

> +	return -ENOTSUPP;
> +}
> +
> +/* must call iput(inode) after this call */
> +static struct inode *inode_from_fd(int ufd, bool check_access)
> +{
> +	struct inode *ret;
> +	struct fd f;
> +	int deny;
> +
> +	f = fdget(ufd);
> +	if (unlikely(!f.file))
> +		return ERR_PTR(-EBADF);
> +	/* TODO?: add this check when called from an eBPF program too (already
> +	* checked by the LSM parent hooks anyway) */
> +	if (unlikely(IS_PRIVATE(file_inode(f.file)))) {
> +		ret = ERR_PTR(-EINVAL);
> +		goto put_fd;
> +	}
> +	/* check if the FD is tied to a mount point */
> +	/* TODO?: add this check when called from an eBPF program too */
> +	if (unlikely(f.file->f_path.mnt->mnt_flags & MNT_INTERNAL)) {
> +		ret = ERR_PTR(-EINVAL);
> +		goto put_fd;
> +	}

a bunch of TODOs do not inspire confidence.

> +	if (check_access) {
> +		/*
> +		* must be allowed to access attributes from this file to then
> +		* be able to compare an inode to its map entry
> +		*/
> +		deny = security_inode_getattr(&f.file->f_path);
> +		if (deny) {
> +			ret = ERR_PTR(deny);
> +			goto put_fd;
> +		}
> +	}
> +	ret = file_inode(f.file);
> +	ihold(ret);
> +
> +put_fd:
> +	fdput(f);
> +	return ret;
> +}
> +
> +/*
> + * The key is a FD when called from a syscall, but an inode address when called
> + * from an eBPF program.
> + */
> +
> +/* called from syscall */
> +int bpf_inode_fd_htab_map_lookup_elem(struct bpf_map *map, int *key, void *value)
> +{
> +	void *ptr;
> +	struct inode *inode;
> +	int ret;
> +
> +	/* check inode access */
> +	inode = inode_from_fd(*key, true);
> +	if (IS_ERR(inode))
> +		return PTR_ERR(inode);
> +
> +	rcu_read_lock();
> +	ptr = htab_map_lookup_elem(map, &inode);
> +	iput(inode);
> +	if (IS_ERR(ptr)) {
> +		ret = PTR_ERR(ptr);
> +	} else if (!ptr) {
> +		ret = -ENOENT;
> +	} else {
> +		ret = 0;
> +		copy_map_value(map, value, ptr);
> +	}
> +	rcu_read_unlock();
> +	return ret;
> +}
> +
> +/* called from kernel */

wrong comment?
kernel side cannot call it, right?

> +int bpf_inode_ptr_locked_htab_map_delete_elem(struct bpf_map *map,
> +		struct inode **key, bool remove_in_inode)
> +{
> +	if (remove_in_inode)
> +		landlock_inode_remove_map(*key, map);
> +	return htab_map_delete_elem(map, key);
> +}
> +
> +/* called from syscall */
> +int bpf_inode_fd_htab_map_delete_elem(struct bpf_map *map, int *key)
> +{
> +	struct inode *inode;
> +	int ret;
> +
> +	/* do not check inode access (similar to directory check) */
> +	inode = inode_from_fd(*key, false);
> +	if (IS_ERR(inode))
> +		return PTR_ERR(inode);
> +	ret = bpf_inode_ptr_locked_htab_map_delete_elem(map, &inode, true);
> +	iput(inode);
> +	return ret;
> +}
> +
> +/* called from syscall */
> +int bpf_inode_fd_htab_map_update_elem(struct bpf_map *map, int *key, void *value,
> +		u64 map_flags)
> +{
> +	struct inode *inode;
> +	int ret;
> +
> +	WARN_ON_ONCE(!rcu_read_lock_held());
> +
> +	/* check inode access */
> +	inode = inode_from_fd(*key, true);
> +	if (IS_ERR(inode))
> +		return PTR_ERR(inode);
> +	ret = htab_map_update_elem(map, &inode, value, map_flags);
> +	if (!ret)
> +		ret = landlock_inode_add_map(inode, map);
> +	iput(inode);
> +	return ret;
> +}
> +
> +static void inode_htab_map_free(struct bpf_map *map)
> +{
> +	struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
> +	struct hlist_nulls_node *n;
> +	struct hlist_nulls_head *head;
> +	struct htab_elem *l;
> +	int i;
> +
> +	for (i = 0; i < htab->n_buckets; i++) {
> +		head = select_bucket(htab, i);
> +		hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
> +			landlock_inode_remove_map(*((struct inode **)l->key), map);
> +		}
> +	}
> +	htab_map_free(map);
> +}

user space can delete the map.
that will trigger inode_htab_map_free() which will call
landlock_inode_remove_map().
which will simply itereate the list and delete from the list.

While in parallel inode can be destoyed and hook_inode_free_security()
will be called.
I think nothing that protects from this race.

> +
> +/*
> + * We need a dedicated helper to deal with inode maps because the key is a
> + * pointer to an opaque data, only provided by the kernel.  This really act
> + * like a (physical or cryptographic) key, which is why it is also not allowed
> + * to get the next key with map_get_next_key().

inode pointer is like cryptographic key? :)

> + */
> +BPF_CALL_2(bpf_inode_map_lookup_elem, struct bpf_map *, map, void *, key)
> +{
> +	WARN_ON_ONCE(!rcu_read_lock_held());
> +	return (unsigned long)htab_map_lookup_elem(map, &key);
> +}
> +
> +const struct bpf_func_proto bpf_inode_map_lookup_elem_proto = {
> +	.func		= bpf_inode_map_lookup_elem,
> +	.gpl_only	= false,
> +	.pkt_access	= true,

pkt_access ? :)

> +	.ret_type	= RET_PTR_TO_MAP_VALUE_OR_NULL,
> +	.arg1_type	= ARG_CONST_MAP_PTR,
> +	.arg2_type	= ARG_PTR_TO_INODE,
> +};
> diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
> index b2a8cb14f28e..e46441c42b68 100644
> --- a/kernel/bpf/syscall.c
> +++ b/kernel/bpf/syscall.c
> @@ -801,6 +801,8 @@ static int map_lookup_elem(union bpf_attr *attr)
>  	} else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
>  		   map->map_type == BPF_MAP_TYPE_STACK) {
>  		err = map->ops->map_peek_elem(map, value);
> +	} else if (map->map_type == BPF_MAP_TYPE_INODE) {
> +		err = bpf_inode_fd_htab_map_lookup_elem(map, key, value);
>  	} else {
>  		rcu_read_lock();
>  		if (map->ops->map_lookup_elem_sys_only)
> @@ -951,6 +953,10 @@ static int map_update_elem(union bpf_attr *attr)
>  	} else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
>  		   map->map_type == BPF_MAP_TYPE_STACK) {
>  		err = map->ops->map_push_elem(map, value, attr->flags);
> +	} else if (map->map_type == BPF_MAP_TYPE_INODE) {
> +		rcu_read_lock();
> +		err = bpf_inode_fd_htab_map_update_elem(map, key, value, attr->flags);
> +		rcu_read_unlock();
>  	} else {
>  		rcu_read_lock();
>  		err = map->ops->map_update_elem(map, key, value, attr->flags);
> @@ -1006,7 +1012,10 @@ static int map_delete_elem(union bpf_attr *attr)
>  	preempt_disable();
>  	__this_cpu_inc(bpf_prog_active);
>  	rcu_read_lock();
> -	err = map->ops->map_delete_elem(map, key);
> +	if (map->map_type == BPF_MAP_TYPE_INODE)
> +		err = bpf_inode_fd_htab_map_delete_elem(map, key);
> +	else
> +		err = map->ops->map_delete_elem(map, key);
>  	rcu_read_unlock();
>  	__this_cpu_dec(bpf_prog_active);
>  	preempt_enable();
> @@ -1018,6 +1027,22 @@ static int map_delete_elem(union bpf_attr *attr)
>  	return err;
>  }
>  
> +int bpf_inode_ptr_unlocked_htab_map_delete_elem(struct bpf_map *map,
> +						struct inode **key, bool remove_in_inode)
> +{
> +	int err;
> +
> +	preempt_disable();
> +	__this_cpu_inc(bpf_prog_active);
> +	rcu_read_lock();
> +	err = bpf_inode_ptr_locked_htab_map_delete_elem(map, key, remove_in_inode);
> +	rcu_read_unlock();
> +	__this_cpu_dec(bpf_prog_active);
> +	preempt_enable();
> +	maybe_wait_bpf_programs(map);

if that function was actually doing synchronize_rcu() the consequences
would have been unpleasant. Fortunately it's a nop in this case.
Please read the code carefully before copy-paste.
Also what do you think the reason of bpf_prog_active above?
What is the reason of rcu_read_lock above?

I think the patch set needs to shrink at least in half to be reviewable.
The way you tie seccomp and lsm is probably the biggest obstacle
than any of the bugs above.
Can you drop seccomp ? and do it as normal lsm ?


^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation for Landlock
  2019-07-21 21:31 ` [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation " Mickaël Salaün
@ 2019-07-31  1:53   ` Randy Dunlap
  2019-08-01 17:03     ` Mickaël Salaün
  0 siblings, 1 reply; 22+ messages in thread
From: Randy Dunlap @ 2019-07-31  1:53 UTC (permalink / raw)
  To: Mickaël Salaün, linux-kernel
  Cc: Alexander Viro, Alexei Starovoitov, Andrew Morton,
	Andy Lutomirski, Arnaldo Carvalho de Melo, Casey Schaufler,
	Daniel Borkmann, David Drysdale, David S . Miller,
	Eric W . Biederman, James Morris, Jann Horn, John Johansen,
	Jonathan Corbet, Kees Cook, Michael Kerrisk,
	Mickaël Salaün, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

On 7/21/19 2:31 PM, Mickaël Salaün wrote:
> This documentation can be built with the Sphinx framework.
> 
> Signed-off-by: Mickaël Salaün <mic@digikod.net>
> Cc: Alexei Starovoitov <ast@kernel.org>
> Cc: Andy Lutomirski <luto@amacapital.net>
> Cc: Daniel Borkmann <daniel@iogearbox.net>
> Cc: David S. Miller <davem@davemloft.net>
> Cc: James Morris <jmorris@namei.org>
> Cc: Jonathan Corbet <corbet@lwn.net>
> Cc: Kees Cook <keescook@chromium.org>
> Cc: Serge E. Hallyn <serge@hallyn.com>
> ---
> 
> Changes since v9:
> * update with expected attach type and expected attach triggers
> 
> Changes since v8:
> * remove documentation related to chaining and tagging according to this
>   patch series
> 
> Changes since v7:
> * update documentation according to the Landlock revamp
> 
> Changes since v6:
> * add a check for ctx->event
> * rename BPF_PROG_TYPE_LANDLOCK to BPF_PROG_TYPE_LANDLOCK_RULE
> * rename Landlock version to ABI to better reflect its purpose and add a
>   dedicated changelog section
> * update tables
> * relax no_new_privs recommendations
> * remove ABILITY_WRITE related functions
> * reword rule "appending" to "prepending" and explain it
> * cosmetic fixes
> 
> Changes since v5:
> * update the rule hierarchy inheritance explanation
> * briefly explain ctx->arg2
> * add ptrace restrictions
> * explain EPERM
> * update example (subtype)
> * use ":manpage:"
> ---
>  Documentation/security/index.rst           |   1 +
>  Documentation/security/landlock/index.rst  |  20 +++
>  Documentation/security/landlock/kernel.rst |  99 ++++++++++++++
>  Documentation/security/landlock/user.rst   | 147 +++++++++++++++++++++
>  4 files changed, 267 insertions(+)
>  create mode 100644 Documentation/security/landlock/index.rst
>  create mode 100644 Documentation/security/landlock/kernel.rst
>  create mode 100644 Documentation/security/landlock/user.rst


> diff --git a/Documentation/security/landlock/kernel.rst b/Documentation/security/landlock/kernel.rst
> new file mode 100644
> index 000000000000..7d1e06d544bf
> --- /dev/null
> +++ b/Documentation/security/landlock/kernel.rst
> @@ -0,0 +1,99 @@
> +==============================
> +Landlock: kernel documentation
> +==============================
> +
> +eBPF properties
> +===============
> +
> +To get an expressive language while still being safe and small, Landlock is
> +based on eBPF. Landlock should be usable by untrusted processes and must
> +therefore expose a minimal attack surface. The eBPF bytecode is minimal,
> +powerful, widely used and designed to be used by untrusted applications. Thus,
> +reusing the eBPF support in the kernel enables a generic approach while
> +minimizing new code.
> +
> +An eBPF program has access to an eBPF context containing some fields used to
> +inspect the current object. These arguments can be used directly (e.g. cookie)
> +or passed to helper functions according to their types (e.g. inode pointer). It
> +is then possible to do complex access checks without race conditions or
> +inconsistent evaluation (i.e.  `incorrect mirroring of the OS code and state
> +<https://www.ndss-symposium.org/ndss2003/traps-and-pitfalls-practical-problems-system-call-interposition-based-security-tools/>`_).
> +
> +A Landlock hook describes a particular access type.  For now, there is two

                                                                 there are two

> +hooks dedicated to filesystem related operations: LANDLOCK_HOOK_FS_PICK and
> +LANDLOCK_HOOK_FS_WALK.  A Landlock program is tied to one hook.  This makes it
> +possible to statically check context accesses, potentially performed by such
> +program, and hence prevents kernel address leaks and ensure the right use of

                                                        ensures

> +hook arguments with eBPF functions.  Any user can add multiple Landlock
> +programs per Landlock hook.  They are stacked and evaluated one after the
> +other, starting from the most recent program, as seccomp-bpf does with its
> +filters.  Underneath, a hook is an abstraction over a set of LSM hooks.
> +
> +
> +Guiding principles
> +==================
> +
> +Unprivileged use
> +----------------
> +
> +* Landlock helpers and context should be usable by any unprivileged and
> +  untrusted program while following the system security policy enforced by
> +  other access control mechanisms (e.g. DAC, LSM).
> +
> +
> +Landlock hook and context
> +-------------------------
> +
> +* A Landlock hook shall be focused on access control on kernel objects instead
> +  of syscall filtering (i.e. syscall arguments), which is the purpose of
> +  seccomp-bpf.
> +* A Landlock context provided by a hook shall express the minimal and more
> +  generic interface to control an access for a kernel object.
> +* A hook shall guaranty that all the BPF function calls from a program are> +  safe.  Thus, the related Landlock context arguments shall always be of the
> +  same type for a particular hook.  For example, a network hook could share
> +  helpers with a file hook because of UNIX socket.  However, the same helpers
> +  may not be compatible for a file system handle and a net handle.
> +* Multiple hooks may use the same context interface.
> +
> +
> +Landlock helpers
> +----------------
> +
> +* Landlock helpers shall be as generic as possible while at the same time being
> +  as simple as possible and following the syscall creation principles (cf.
> +  *Documentation/adding-syscalls.txt*).
> +* The only behavior change allowed on a helper is to fix a (logical) bug to
> +  match the initial semantic.
> +* Helpers shall be reentrant, i.e. only take inputs from arguments (e.g. from
> +  the BPF context), to enable a hook to use a cache.  Future program options
> +  might change this cache behavior.
> +* It is quite easy to add new helpers to extend Landlock.  The main concern
> +  should be about the possibility to leak information from the kernel that may
> +  not be accessible otherwise (i.e. side-channel attack).
> +
> +
> +Questions and answers
> +=====================
> +
> +Why not create a custom hook for each kind of action?
> +-----------------------------------------------------
> +
> +Landlock programs can handle these checks.  Adding more exceptions to the
> +kernel code would lead to more code complexity.  A decision to ignore a kind of
> +action can and should be done at the beginning of a Landlock program.
> +
> +
> +Why a program does not return an errno or a kill code?
> +------------------------------------------------------
> +
> +seccomp filters can return multiple kind of code, including an errno value or a

                                       kinds

> +kill signal, which may be convenient for access control.  Those return codes
> +are hardwired in the userland ABI.  Instead, Landlock's approach is to return a
> +boolean to allow or deny an action, which is much simpler and more generic.
> +Moreover, we do not really have a choice because, unlike to seccomp, Landlock
> +programs are not enforced at the syscall entry point but may be executed at any
> +point in the kernel (through LSM hooks) where an errno return code may not make
> +sense.  However, with this simple ABI and with the ability to call helpers,
> +Landlock may gain features similar to seccomp-bpf in the future while being
> +compatible with previous programs.
> diff --git a/Documentation/security/landlock/user.rst b/Documentation/security/landlock/user.rst
> new file mode 100644
> index 000000000000..14c4f3b377bd
> --- /dev/null
> +++ b/Documentation/security/landlock/user.rst
> @@ -0,0 +1,147 @@
> +================================
> +Landlock: userland documentation
> +================================
> +
> +Landlock programs
> +=================
> +
> +eBPF programs are used to create security programs.  They are contained and can
> +call only a whitelist of dedicated functions. Moreover, they can only loop
> +under strict conditions, which protects from denial of service.  More
> +information on BPF can be found in *Documentation/networking/filter.txt*.
> +
> +
> +Writing a program
> +-----------------
> +
> +To enforce a security policy, a thread first needs to create a Landlock program.
> +The easiest way to write an eBPF program depicting a security program is to write
> +it in the C language.  As described in *samples/bpf/README.rst*, LLVM can
> +compile such programs.  Files *samples/bpf/landlock1_kern.c* and those in
> +*tools/testing/selftests/landlock/* can be used as examples.
> +
> +Once the eBPF program is created, the next step is to create the metadata
> +describing the Landlock program.  This metadata includes an expected attach type which
> +contains the hook type to which the program is tied, and expected attach
> +triggers which identify the actions for which the program should be run.
> +
> +A hook is a policy decision point which exposes the same context type for
> +each program evaluation.
> +
> +A Landlock hook describes the kind of kernel object for which a program will be
> +triggered to allow or deny an action.  For example, the hook
> +BPF_LANDLOCK_FS_PICK can be triggered every time a landlocked thread performs a
> +set of action related to the filesystem (e.g. open, read, write, mount...).

          actions

> +This actions are identified by the `triggers` bitfield.
> +
> +The next step is to fill a :c:type:`struct bpf_load_program_attr
> +<bpf_load_program_attr>` with BPF_PROG_TYPE_LANDLOCK_HOOK, the expected attach
> +type and other BPF program metadata.  This bpf_attr must then be passed to the
> +:manpage:`bpf(2)` syscall alongside the BPF_PROG_LOAD command.  If everything
> +is deemed correct by the kernel, the thread gets a file descriptor referring to
> +this program.
> +
> +In the following code, the *insn* variable is an array of BPF instructions
> +which can be extracted from an ELF file as is done in bpf_load_file() from
> +*samples/bpf/bpf_load.c*.

A little confusing.  Is there a mixup of <insn> and <insns>?

> +
> +.. code-block:: c
> +
> +    int prog_fd;
> +    struct bpf_load_program_attr load_attr;
> +
> +    memset(&load_attr, 0, sizeof(struct bpf_load_program_attr));
> +    load_attr.prog_type = BPF_PROG_TYPE_LANDLOCK_HOOK;
> +    load_attr.expected_attach_type = BPF_LANDLOCK_FS_PICK;
> +    load_attr.expected_attach_triggers = LANDLOCK_TRIGGER_FS_PICK_OPEN;
> +    load_attr.insns = insns;
> +    load_attr.insns_cnt = sizeof(insn) / sizeof(struct bpf_insn);
> +    load_attr.license = "GPL";
> +
> +    prog_fd = bpf_load_program_xattr(&load_attr, log_buf, log_buf_sz);
> +    if (prog_fd == -1)
> +        exit(1);
> +
> +
> +Enforcing a program
> +-------------------
> +
> +Once the Landlock program has been created or received (e.g. through a UNIX
> +socket), the thread willing to sandbox itself (and its future children) should
> +perform the following two steps.
> +
> +The thread should first request to never be allowed to get new privileges with a
> +call to :manpage:`prctl(2)` and the PR_SET_NO_NEW_PRIVS option.  More
> +information can be found in *Documentation/prctl/no_new_privs.txt*.
> +
> +.. code-block:: c
> +
> +    if (prctl(PR_SET_NO_NEW_PRIVS, 1, NULL, 0, 0))
> +        exit(1);
> +
> +A thread can apply a program to itself by using the :manpage:`seccomp(2)` syscall.
> +The operation is SECCOMP_PREPEND_LANDLOCK_PROG, the flags must be empty and the
> +*args* argument must point to a valid Landlock program file descriptor.
> +
> +.. code-block:: c
> +
> +    if (seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, &fd))
> +        exit(1);
> +
> +If the syscall succeeds, the program is now enforced on the calling thread and
> +will be enforced on all its subsequently created children of the thread as
> +well.  Once a thread is landlocked, there is no way to remove this security
> +policy, only stacking more restrictions is allowed.  The program evaluation is
> +performed from the newest to the oldest.
> +
> +When a syscall ask for an action on a kernel object, if this action is denied,

                  asks

> +then an EACCES errno code is returned through the syscall.
> +
> +
> +.. _inherited_programs:
> +
> +Inherited programs
> +------------------
> +
> +Every new thread resulting from a :manpage:`clone(2)` inherits Landlock program
> +restrictions from its parent.  This is similar to the seccomp inheritance as
> +described in *Documentation/prctl/seccomp_filter.txt*.
> +
> +
> +Ptrace restrictions
> +-------------------
> +
> +A landlocked process has less privileges than a non-landlocked process and must
> +then be subject to additional restrictions when manipulating another process.
> +To be allowed to use :manpage:`ptrace(2)` and related syscalls on a target
> +process, a landlocked process must have a subset of the target process programs.
            ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Maybe that last statement is correct, but it seems to me that it is missing something.

> +
> +
> +Landlock structures and constants
> +=================================
> +
> +Hook types
> +----------
> +
> +.. kernel-doc:: include/uapi/linux/landlock.h
> +    :functions: landlock_hook_type
> +
> +
> +Contexts
> +--------
> +
> +.. kernel-doc:: include/uapi/linux/landlock.h
> +    :functions: landlock_ctx_fs_pick landlock_ctx_fs_walk landlock_ctx_fs_get
> +
> +
> +Triggers for fs_pick
> +--------------------
> +
> +.. kernel-doc:: include/uapi/linux/landlock.h
> +    :functions: landlock_triggers
> +
> +
> +Additional documentation
> +========================
> +
> +See https://landlock.io
> 


-- 
~Randy

^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-07-27  1:40   ` Alexei Starovoitov
@ 2019-07-31 18:46     ` Mickaël Salaün
  2019-07-31 18:58       ` Alexei Starovoitov
  2019-09-08 22:09       ` Mickaël Salaün
  0 siblings, 2 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-31 18:46 UTC (permalink / raw)
  To: Alexei Starovoitov, Mickaël Salaün
  Cc: linux-kernel, Alexander Viro, Alexei Starovoitov, Andrew Morton,
	Andy Lutomirski, Arnaldo Carvalho de Melo, Casey Schaufler,
	Daniel Borkmann, David Drysdale, David S . Miller,
	Eric W . Biederman, James Morris, Jann Horn, John Johansen,
	Jonathan Corbet, Kees Cook, Michael Kerrisk, Paul Moore,
	Sargun Dhillon, Serge E . Hallyn, Shuah Khan, Stephen Smalley,
	Tejun Heo, Tetsuo Handa, Thomas Graf, Tycho Andersen,
	Will Drewry, kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev


On 27/07/2019 03:40, Alexei Starovoitov wrote:
> On Sun, Jul 21, 2019 at 11:31:12PM +0200, Mickaël Salaün wrote:
>> FIXME: 64-bits in the doc

FYI, this FIXME was fixed, just not removed from this message. :)

>>
>> This new map store arbitrary values referenced by inode keys.  The map
>> can be updated from user space with file descriptor pointing to inodes
>> tied to a file system.  From an eBPF (Landlock) program point of view,
>> such a map is read-only and can only be used to retrieved a value tied
>> to a given inode.  This is useful to recognize an inode tagged by user
>> space, without access right to this inode (i.e. no need to have a write
>> access to this inode).
>>
>> Add dedicated BPF functions to handle this type of map:
>> * bpf_inode_htab_map_update_elem()
>> * bpf_inode_htab_map_lookup_elem()
>> * bpf_inode_htab_map_delete_elem()
>>
>> This new map require a dedicated helper inode_map_lookup_elem() because
>> of the key which is a pointer to an opaque data (only provided by the
>> kernel).  This act like a (physical or cryptographic) key, which is why
>> it is also not allowed to get the next key.
>>
>> Signed-off-by: Mickaël Salaün <mic@digikod.net>
>
> there are too many things to comment on.
> Let's do this patch.
>
> imo inode_map concept is interesting, but see below...
>
>> +
>> +    /*
>> +     * Limit number of entries in an inode map to the maximum number of
>> +     * open files for the current process. The maximum number of file
>> +     * references (including all inode maps) for a process is then
>> +     * (RLIMIT_NOFILE - 1) * RLIMIT_NOFILE. If the process' RLIMIT_NOFILE
>> +     * is 0, then any entry update is forbidden.
>> +     *
>> +     * An eBPF program can inherit all the inode map FD. The worse case is
>> +     * to fill a bunch of arraymaps, create an eBPF program, close the
>> +     * inode map FDs, and start again. The maximum number of inode map
>> +     * entries can then be close to RLIMIT_NOFILE^3.
>> +     */
>> +    if (attr->max_entries > rlimit(RLIMIT_NOFILE))
>> +            return -EMFILE;
>
> rlimit is checked, but no fd are consumed.
> Once created such inode map_fd can be passed to a different process.
> map_fd can be pinned into bpffs.
> etc.
> what the value of the check?

I was looking for the most meaningful limit for a process, and rlimit is
the best I found. As the limit of open FD per processes, rlimit is not
perfect, but I think the semantic is close here (e.g. a process can also
pass FD through unix socket).

>
>> +
>> +    /* decorelate UAPI from kernel API */
>> +    attr->key_size = sizeof(struct inode *);
>> +
>> +    return htab_map_alloc_check(attr);
>> +}
>> +
>> +static void inode_htab_put_key(void *key)
>> +{
>> +    struct inode **inode = key;
>> +
>> +    if ((*inode)->i_state & I_FREEING)
>> +            return;
>
> checking the state without take a lock? isn't it racy?

This should only trigger when called from security_inode_free(). I'll
add a comment.

>
>> +    iput(*inode);
>> +}
>> +
>> +/* called from syscall or (never) from eBPF program */
>> +static int map_get_next_no_key(struct bpf_map *map, void *key, void *next_key)
>> +{
>> +    /* do not leak a file descriptor */
>
> what this comment suppose to mean?

Because a key is a reference to an inode, a possible return value for
this function could be a file descriptor pointing to this inode (the
same way a file descriptor is use to add an element). For now, I don't
want to implement a way for a process with such a map to extract such
inode, which I compare to a possible leak (of information, not kernel
memory nor object). This could be implemented in the future if there is
value in it (and probably some additional safeguards), though.

>
>> +    return -ENOTSUPP;
>> +}
>> +
>> +/* must call iput(inode) after this call */
>> +static struct inode *inode_from_fd(int ufd, bool check_access)
>> +{
>> +    struct inode *ret;
>> +    struct fd f;
>> +    int deny;
>> +
>> +    f = fdget(ufd);
>> +    if (unlikely(!f.file))
>> +            return ERR_PTR(-EBADF);
>> +    /* TODO?: add this check when called from an eBPF program too (already
>> +    * checked by the LSM parent hooks anyway) */
>> +    if (unlikely(IS_PRIVATE(file_inode(f.file)))) {
>> +            ret = ERR_PTR(-EINVAL);
>> +            goto put_fd;
>> +    }
>> +    /* check if the FD is tied to a mount point */
>> +    /* TODO?: add this check when called from an eBPF program too */
>> +    if (unlikely(f.file->f_path.mnt->mnt_flags & MNT_INTERNAL)) {
>> +            ret = ERR_PTR(-EINVAL);
>> +            goto put_fd;
>> +    }
>
> a bunch of TODOs do not inspire confidence.

I think the current implement is good, but these TODOs are here to draw
attention on particular points for which I would like external review
and opinion (hence the "?").

>
>> +    if (check_access) {
>> +            /*
>> +            * must be allowed to access attributes from this file to then
>> +            * be able to compare an inode to its map entry
>> +            */
>> +            deny = security_inode_getattr(&f.file->f_path);
>> +            if (deny) {
>> +                    ret = ERR_PTR(deny);
>> +                    goto put_fd;
>> +            }
>> +    }
>> +    ret = file_inode(f.file);
>> +    ihold(ret);
>> +
>> +put_fd:
>> +    fdput(f);
>> +    return ret;
>> +}
>> +
>> +/*
>> + * The key is a FD when called from a syscall, but an inode address when called
>> + * from an eBPF program.
>> + */
>> +
>> +/* called from syscall */
>> +int bpf_inode_fd_htab_map_lookup_elem(struct bpf_map *map, int *key, void *value)
>> +{
>> +    void *ptr;
>> +    struct inode *inode;
>> +    int ret;
>> +
>> +    /* check inode access */
>> +    inode = inode_from_fd(*key, true);
>> +    if (IS_ERR(inode))
>> +            return PTR_ERR(inode);
>> +
>> +    rcu_read_lock();
>> +    ptr = htab_map_lookup_elem(map, &inode);
>> +    iput(inode);
>> +    if (IS_ERR(ptr)) {
>> +            ret = PTR_ERR(ptr);
>> +    } else if (!ptr) {
>> +            ret = -ENOENT;
>> +    } else {
>> +            ret = 0;
>> +            copy_map_value(map, value, ptr);
>> +    }
>> +    rcu_read_unlock();
>> +    return ret;
>> +}
>> +
>> +/* called from kernel */
>
> wrong comment?
> kernel side cannot call it, right?

This is called from bpf_inode_fd_htab_map_delete_elem() (code just
beneath), and from
kernel/bpf/syscall.c:bpf_inode_ptr_unlocked_htab_map_delet_elem() which
can be called by security_inode_free() (hook_inode_free_security).

>
>> +int bpf_inode_ptr_locked_htab_map_delete_elem(struct bpf_map *map,
>> +            struct inode **key, bool remove_in_inode)
>> +{
>> +    if (remove_in_inode)
>> +            landlock_inode_remove_map(*key, map);
>> +    return htab_map_delete_elem(map, key);
>> +}
>> +
>> +/* called from syscall */
>> +int bpf_inode_fd_htab_map_delete_elem(struct bpf_map *map, int *key)
>> +{
>> +    struct inode *inode;
>> +    int ret;
>> +
>> +    /* do not check inode access (similar to directory check) */
>> +    inode = inode_from_fd(*key, false);
>> +    if (IS_ERR(inode))
>> +            return PTR_ERR(inode);
>> +    ret = bpf_inode_ptr_locked_htab_map_delete_elem(map, &inode, true);
>> +    iput(inode);
>> +    return ret;
>> +}
>> +
>> +/* called from syscall */
>> +int bpf_inode_fd_htab_map_update_elem(struct bpf_map *map, int *key, void *value,
>> +            u64 map_flags)
>> +{
>> +    struct inode *inode;
>> +    int ret;
>> +
>> +    WARN_ON_ONCE(!rcu_read_lock_held());
>> +
>> +    /* check inode access */
>> +    inode = inode_from_fd(*key, true);
>> +    if (IS_ERR(inode))
>> +            return PTR_ERR(inode);
>> +    ret = htab_map_update_elem(map, &inode, value, map_flags);
>> +    if (!ret)
>> +            ret = landlock_inode_add_map(inode, map);
>> +    iput(inode);
>> +    return ret;
>> +}
>> +
>> +static void inode_htab_map_free(struct bpf_map *map)
>> +{
>> +    struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
>> +    struct hlist_nulls_node *n;
>> +    struct hlist_nulls_head *head;
>> +    struct htab_elem *l;
>> +    int i;
>> +
>> +    for (i = 0; i < htab->n_buckets; i++) {
>> +            head = select_bucket(htab, i);
>> +            hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
>> +                    landlock_inode_remove_map(*((struct inode **)l->key), map);
>> +            }
>> +    }
>> +    htab_map_free(map);
>> +}
>
> user space can delete the map.
> that will trigger inode_htab_map_free() which will call
> landlock_inode_remove_map().
> which will simply itereate the list and delete from the list.

landlock_inode_remove_map() removes the reference to the map (being
freed) from the inode (with an RCU lock).

>
> While in parallel inode can be destoyed and hook_inode_free_security()
> will be called.
> I think nothing that protects from this race.

According to security_inode_free(), the inode is effectively freed after
the RCU grace period. However, I forgot to call bpf_map_inc() in
landlock_inode_add_map(), which would prevent the map to be freed
outside of the security_inode_free(). I'll fix that.

>
>> +
>> +/*
>> + * We need a dedicated helper to deal with inode maps because the key is a
>> + * pointer to an opaque data, only provided by the kernel.  This really act
>> + * like a (physical or cryptographic) key, which is why it is also not allowed
>> + * to get the next key with map_get_next_key().
>
> inode pointer is like cryptographic key? :)

I wanted to highlight the fact that, contrary to other map key types,
the value of this one should not be readable, only usable. A "secret
value" is more appropriate but still confusing. I'll rephrase that.

>
>> + */
>> +BPF_CALL_2(bpf_inode_map_lookup_elem, struct bpf_map *, map, void *, key)
>> +{
>> +    WARN_ON_ONCE(!rcu_read_lock_held());
>> +    return (unsigned long)htab_map_lookup_elem(map, &key);
>> +}
>> +
>> +const struct bpf_func_proto bpf_inode_map_lookup_elem_proto = {
>> +    .func           = bpf_inode_map_lookup_elem,
>> +    .gpl_only       = false,
>> +    .pkt_access     = true,
>
> pkt_access ? :)

This slipped in with this rebase, I'll remove it. :)

>
>> +    .ret_type       = RET_PTR_TO_MAP_VALUE_OR_NULL,
>> +    .arg1_type      = ARG_CONST_MAP_PTR,
>> +    .arg2_type      = ARG_PTR_TO_INODE,
>> +};
>> diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
>> index b2a8cb14f28e..e46441c42b68 100644
>> --- a/kernel/bpf/syscall.c
>> +++ b/kernel/bpf/syscall.c
>> @@ -801,6 +801,8 @@ static int map_lookup_elem(union bpf_attr *attr)
>>      } else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
>>                 map->map_type == BPF_MAP_TYPE_STACK) {
>>              err = map->ops->map_peek_elem(map, value);
>> +    } else if (map->map_type == BPF_MAP_TYPE_INODE) {
>> +            err = bpf_inode_fd_htab_map_lookup_elem(map, key, value);
>>      } else {
>>              rcu_read_lock();
>>              if (map->ops->map_lookup_elem_sys_only)
>> @@ -951,6 +953,10 @@ static int map_update_elem(union bpf_attr *attr)
>>      } else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
>>                 map->map_type == BPF_MAP_TYPE_STACK) {
>>              err = map->ops->map_push_elem(map, value, attr->flags);
>> +    } else if (map->map_type == BPF_MAP_TYPE_INODE) {
>> +            rcu_read_lock();
>> +            err = bpf_inode_fd_htab_map_update_elem(map, key, value, attr->flags);
>> +            rcu_read_unlock();
>>      } else {
>>              rcu_read_lock();
>>              err = map->ops->map_update_elem(map, key, value, attr->flags);
>> @@ -1006,7 +1012,10 @@ static int map_delete_elem(union bpf_attr *attr)
>>      preempt_disable();
>>      __this_cpu_inc(bpf_prog_active);
>>      rcu_read_lock();
>> -    err = map->ops->map_delete_elem(map, key);
>> +    if (map->map_type == BPF_MAP_TYPE_INODE)
>> +            err = bpf_inode_fd_htab_map_delete_elem(map, key);
>> +    else
>> +            err = map->ops->map_delete_elem(map, key);
>>      rcu_read_unlock();
>>      __this_cpu_dec(bpf_prog_active);
>>      preempt_enable();
>> @@ -1018,6 +1027,22 @@ static int map_delete_elem(union bpf_attr *attr)
>>      return err;
>>  }
>>
>> +int bpf_inode_ptr_unlocked_htab_map_delete_elem(struct bpf_map *map,
>> +                                            struct inode **key, bool remove_in_inode)
>> +{
>> +    int err;
>> +
>> +    preempt_disable();
>> +    __this_cpu_inc(bpf_prog_active);
>> +    rcu_read_lock();
>> +    err = bpf_inode_ptr_locked_htab_map_delete_elem(map, key, remove_in_inode);
>> +    rcu_read_unlock();
>> +    __this_cpu_dec(bpf_prog_active);
>> +    preempt_enable();
>> +    maybe_wait_bpf_programs(map);
>
> if that function was actually doing synchronize_rcu() the consequences
> would have been unpleasant. Fortunately it's a nop in this case.
> Please read the code carefully before copy-paste.
> Also what do you think the reason of bpf_prog_active above?
> What is the reason of rcu_read_lock above?

The RCU is used as for every map modifications (usually from userspace).
I wasn't sure about the other protections so I kept the same (generic)
checks as in map_delete_elem() (just above) because this function follow
the same semantic. What can I safely remove?

>
> I think the patch set needs to shrink at least in half to be reviewable.
> The way you tie seccomp and lsm is probably the biggest obstacle
> than any of the bugs above.
> Can you drop seccomp ? and do it as normal lsm ?

The seccomp/enforcement part is needed to have a minimum viable product,
i.e. a process able to sandbox itself. Are you suggesting to first merge
a version when it is only possible to create inode maps but not use them
in an useful way (i.e. for sandboxing)? I can do it if it's OK with you,
and I hope it will not be a problem for the security folks if it can
help to move forward.

--
Mickaël Salaün
ANSSI/SDE/ST/LAM

Les données à caractère personnel recueillies et traitées dans le cadre de cet échange, le sont à seule fin d’exécution d’une relation professionnelle et s’opèrent dans cette seule finalité et pour la durée nécessaire à cette relation. Si vous souhaitez faire usage de vos droits de consultation, de rectification et de suppression de vos données, veuillez contacter contact.rgpd@sgdsn.gouv.fr. Si vous avez reçu ce message par erreur, nous vous remercions d’en informer l’expéditeur et de détruire le message. The personal data collected and processed during this exchange aims solely at completing a business relationship and is limited to the necessary duration of that relationship. If you wish to use your rights of consultation, rectification and deletion of your data, please contact: contact.rgpd@sgdsn.gouv.fr. If you have received this message in error, we thank you for informing the sender and destroying the message.

^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-07-31 18:46     ` Mickaël Salaün
@ 2019-07-31 18:58       ` Alexei Starovoitov
  2019-07-31 19:11         ` Mickaël Salaün
  2019-09-08 22:09       ` Mickaël Salaün
  1 sibling, 1 reply; 22+ messages in thread
From: Alexei Starovoitov @ 2019-07-31 18:58 UTC (permalink / raw)
  To: Mickaël Salaün
  Cc: Mickaël Salaün, LKML, Alexander Viro,
	Alexei Starovoitov, Andrew Morton, Andy Lutomirski,
	Arnaldo Carvalho de Melo, Casey Schaufler, Daniel Borkmann,
	David Drysdale, David S . Miller, Eric W . Biederman,
	James Morris, Jann Horn, John Johansen, Jonathan Corbet,
	Kees Cook, Michael Kerrisk, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	Kernel Hardening, Linux API, Linux-Fsdevel, LSM List,
	Network Development

On Wed, Jul 31, 2019 at 11:46 AM Mickaël Salaün
<mickael.salaun@ssi.gouv.fr> wrote:
> >> +    for (i = 0; i < htab->n_buckets; i++) {
> >> +            head = select_bucket(htab, i);
> >> +            hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
> >> +                    landlock_inode_remove_map(*((struct inode **)l->key), map);
> >> +            }
> >> +    }
> >> +    htab_map_free(map);
> >> +}
> >
> > user space can delete the map.
> > that will trigger inode_htab_map_free() which will call
> > landlock_inode_remove_map().
> > which will simply itereate the list and delete from the list.
>
> landlock_inode_remove_map() removes the reference to the map (being
> freed) from the inode (with an RCU lock).

I'm going to ignore everything else for now and focus only on this bit,
since it's fundamental issue to address before this discussion can
go any further.
rcu_lock is not a spin_lock. I'm pretty sure you know this.
But you're arguing that it's somehow protecting from the race
I mentioned above?

^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-07-31 18:58       ` Alexei Starovoitov
@ 2019-07-31 19:11         ` Mickaël Salaün
  2019-08-01 17:35           ` Alexei Starovoitov
  0 siblings, 1 reply; 22+ messages in thread
From: Mickaël Salaün @ 2019-07-31 19:11 UTC (permalink / raw)
  To: Alexei Starovoitov
  Cc: Mickaël Salaün, LKML, Alexander Viro,
	Alexei Starovoitov, Andrew Morton, Andy Lutomirski,
	Arnaldo Carvalho de Melo, Casey Schaufler, Daniel Borkmann,
	David Drysdale, David S . Miller, Eric W . Biederman,
	James Morris, Jann Horn, John Johansen, Jonathan Corbet,
	Kees Cook, Michael Kerrisk, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	Kernel Hardening, Linux API, Linux-Fsdevel, LSM List,
	Network Development



On 31/07/2019 20:58, Alexei Starovoitov wrote:
> On Wed, Jul 31, 2019 at 11:46 AM Mickaël Salaün
> <mickael.salaun@ssi.gouv.fr> wrote:
>>>> +    for (i = 0; i < htab->n_buckets; i++) {
>>>> +            head = select_bucket(htab, i);
>>>> +            hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
>>>> +                    landlock_inode_remove_map(*((struct inode **)l->key), map);
>>>> +            }
>>>> +    }
>>>> +    htab_map_free(map);
>>>> +}
>>>
>>> user space can delete the map.
>>> that will trigger inode_htab_map_free() which will call
>>> landlock_inode_remove_map().
>>> which will simply itereate the list and delete from the list.
>>
>> landlock_inode_remove_map() removes the reference to the map (being
>> freed) from the inode (with an RCU lock).
>
> I'm going to ignore everything else for now and focus only on this bit,
> since it's fundamental issue to address before this discussion can
> go any further.
> rcu_lock is not a spin_lock. I'm pretty sure you know this.
> But you're arguing that it's somehow protecting from the race
> I mentioned above?
>

I was just clarifying your comment to avoid misunderstanding about what
is being removed.

As said in the full response, there is currently a race but, if I add a
bpf_map_inc() call when the map is referenced by inode->security, then I
don't see how a race could occur because such added map could only be
freed in a security_inode_free() (as long as it retains a reference to
this inode).


--
Mickaël Salaün
ANSSI/SDE/ST/LAM

Les données à caractère personnel recueillies et traitées dans le cadre de cet échange, le sont à seule fin d’exécution d’une relation professionnelle et s’opèrent dans cette seule finalité et pour la durée nécessaire à cette relation. Si vous souhaitez faire usage de vos droits de consultation, de rectification et de suppression de vos données, veuillez contacter contact.rgpd@sgdsn.gouv.fr. Si vous avez reçu ce message par erreur, nous vous remercions d’en informer l’expéditeur et de détruire le message. The personal data collected and processed during this exchange aims solely at completing a business relationship and is limited to the necessary duration of that relationship. If you wish to use your rights of consultation, rectification and deletion of your data, please contact: contact.rgpd@sgdsn.gouv.fr. If you have received this message in error, we thank you for informing the sender and destroying the message.

^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation for Landlock
  2019-07-31  1:53   ` Randy Dunlap
@ 2019-08-01 17:03     ` Mickaël Salaün
  2019-08-01 17:49       ` Randy Dunlap
  0 siblings, 1 reply; 22+ messages in thread
From: Mickaël Salaün @ 2019-08-01 17:03 UTC (permalink / raw)
  To: Randy Dunlap, Mickaël Salaün, linux-kernel
  Cc: Alexander Viro, Alexei Starovoitov, Andrew Morton,
	Andy Lutomirski, Arnaldo Carvalho de Melo, Casey Schaufler,
	Daniel Borkmann, David Drysdale, David S . Miller,
	Eric W . Biederman, James Morris, Jann Horn, John Johansen,
	Jonathan Corbet, Kees Cook, Michael Kerrisk, Paul Moore,
	Sargun Dhillon, Serge E . Hallyn, Shuah Khan, Stephen Smalley,
	Tejun Heo, Tetsuo Handa, Thomas Graf, Tycho Andersen,
	Will Drewry, kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

Thanks for this spelling fixes. Some comments:

On 31/07/2019 03:53, Randy Dunlap wrote:
> On 7/21/19 2:31 PM, Mickaël Salaün wrote:
>> This documentation can be built with the Sphinx framework.
>>
>> Signed-off-by: Mickaël Salaün <mic@digikod.net>
>> Cc: Alexei Starovoitov <ast@kernel.org>
>> Cc: Andy Lutomirski <luto@amacapital.net>
>> Cc: Daniel Borkmann <daniel@iogearbox.net>
>> Cc: David S. Miller <davem@davemloft.net>
>> Cc: James Morris <jmorris@namei.org>
>> Cc: Jonathan Corbet <corbet@lwn.net>
>> Cc: Kees Cook <keescook@chromium.org>
>> Cc: Serge E. Hallyn <serge@hallyn.com>
>> ---
>>
>> Changes since v9:
>> * update with expected attach type and expected attach triggers
>>
>> Changes since v8:
>> * remove documentation related to chaining and tagging according to this
>>   patch series
>>
>> Changes since v7:
>> * update documentation according to the Landlock revamp
>>
>> Changes since v6:
>> * add a check for ctx->event
>> * rename BPF_PROG_TYPE_LANDLOCK to BPF_PROG_TYPE_LANDLOCK_RULE
>> * rename Landlock version to ABI to better reflect its purpose and add a
>>   dedicated changelog section
>> * update tables
>> * relax no_new_privs recommendations
>> * remove ABILITY_WRITE related functions
>> * reword rule "appending" to "prepending" and explain it
>> * cosmetic fixes
>>
>> Changes since v5:
>> * update the rule hierarchy inheritance explanation
>> * briefly explain ctx->arg2
>> * add ptrace restrictions
>> * explain EPERM
>> * update example (subtype)
>> * use ":manpage:"
>> ---
>>  Documentation/security/index.rst           |   1 +
>>  Documentation/security/landlock/index.rst  |  20 +++
>>  Documentation/security/landlock/kernel.rst |  99 ++++++++++++++
>>  Documentation/security/landlock/user.rst   | 147 +++++++++++++++++++++
>>  4 files changed, 267 insertions(+)
>>  create mode 100644 Documentation/security/landlock/index.rst
>>  create mode 100644 Documentation/security/landlock/kernel.rst
>>  create mode 100644 Documentation/security/landlock/user.rst
>
>
>> diff --git a/Documentation/security/landlock/kernel.rst b/Documentation/security/landlock/kernel.rst
>> new file mode 100644
>> index 000000000000..7d1e06d544bf
>> --- /dev/null
>> +++ b/Documentation/security/landlock/kernel.rst
>> @@ -0,0 +1,99 @@
>> +==============================
>> +Landlock: kernel documentation
>> +==============================
>> +
>> +eBPF properties
>> +===============
>> +
>> +To get an expressive language while still being safe and small, Landlock is
>> +based on eBPF. Landlock should be usable by untrusted processes and must
>> +therefore expose a minimal attack surface. The eBPF bytecode is minimal,
>> +powerful, widely used and designed to be used by untrusted applications. Thus,
>> +reusing the eBPF support in the kernel enables a generic approach while
>> +minimizing new code.
>> +
>> +An eBPF program has access to an eBPF context containing some fields used to
>> +inspect the current object. These arguments can be used directly (e.g. cookie)
>> +or passed to helper functions according to their types (e.g. inode pointer). It
>> +is then possible to do complex access checks without race conditions or
>> +inconsistent evaluation (i.e.  `incorrect mirroring of the OS code and state
>> +<https://www.ndss-symposium.org/ndss2003/traps-and-pitfalls-practical-problems-system-call-interposition-based-security-tools/>`_).
>> +
>> +A Landlock hook describes a particular access type.  For now, there is two
>
>                                                                  there are two
>
>> +hooks dedicated to filesystem related operations: LANDLOCK_HOOK_FS_PICK and
>> +LANDLOCK_HOOK_FS_WALK.  A Landlock program is tied to one hook.  This makes it
>> +possible to statically check context accesses, potentially performed by such
>> +program, and hence prevents kernel address leaks and ensure the right use of
>
>                                                         ensures
>
>> +hook arguments with eBPF functions.  Any user can add multiple Landlock
>> +programs per Landlock hook.  They are stacked and evaluated one after the
>> +other, starting from the most recent program, as seccomp-bpf does with its
>> +filters.  Underneath, a hook is an abstraction over a set of LSM hooks.
>> +
>> +
>> +Guiding principles
>> +==================
>> +
>> +Unprivileged use
>> +----------------
>> +
>> +* Landlock helpers and context should be usable by any unprivileged and
>> +  untrusted program while following the system security policy enforced by
>> +  other access control mechanisms (e.g. DAC, LSM).
>> +
>> +
>> +Landlock hook and context
>> +-------------------------
>> +
>> +* A Landlock hook shall be focused on access control on kernel objects instead
>> +  of syscall filtering (i.e. syscall arguments), which is the purpose of
>> +  seccomp-bpf.
>> +* A Landlock context provided by a hook shall express the minimal and more
>> +  generic interface to control an access for a kernel object.
>> +* A hook shall guaranty that all the BPF function calls from a program are> +  safe.  Thus, the related Landlock context arguments shall always be of the
>> +  same type for a particular hook.  For example, a network hook could share
>> +  helpers with a file hook because of UNIX socket.  However, the same helpers
>> +  may not be compatible for a file system handle and a net handle.
>> +* Multiple hooks may use the same context interface.
>> +
>> +
>> +Landlock helpers
>> +----------------
>> +
>> +* Landlock helpers shall be as generic as possible while at the same time being
>> +  as simple as possible and following the syscall creation principles (cf.
>> +  *Documentation/adding-syscalls.txt*).
>> +* The only behavior change allowed on a helper is to fix a (logical) bug to
>> +  match the initial semantic.
>> +* Helpers shall be reentrant, i.e. only take inputs from arguments (e.g. from
>> +  the BPF context), to enable a hook to use a cache.  Future program options
>> +  might change this cache behavior.
>> +* It is quite easy to add new helpers to extend Landlock.  The main concern
>> +  should be about the possibility to leak information from the kernel that may
>> +  not be accessible otherwise (i.e. side-channel attack).
>> +
>> +
>> +Questions and answers
>> +=====================
>> +
>> +Why not create a custom hook for each kind of action?
>> +-----------------------------------------------------
>> +
>> +Landlock programs can handle these checks.  Adding more exceptions to the
>> +kernel code would lead to more code complexity.  A decision to ignore a kind of
>> +action can and should be done at the beginning of a Landlock program.
>> +
>> +
>> +Why a program does not return an errno or a kill code?
>> +------------------------------------------------------
>> +
>> +seccomp filters can return multiple kind of code, including an errno value or a
>
>                                        kinds
>
>> +kill signal, which may be convenient for access control.  Those return codes
>> +are hardwired in the userland ABI.  Instead, Landlock's approach is to return a
>> +boolean to allow or deny an action, which is much simpler and more generic.
>> +Moreover, we do not really have a choice because, unlike to seccomp, Landlock
>> +programs are not enforced at the syscall entry point but may be executed at any
>> +point in the kernel (through LSM hooks) where an errno return code may not make
>> +sense.  However, with this simple ABI and with the ability to call helpers,
>> +Landlock may gain features similar to seccomp-bpf in the future while being
>> +compatible with previous programs.
>> diff --git a/Documentation/security/landlock/user.rst b/Documentation/security/landlock/user.rst
>> new file mode 100644
>> index 000000000000..14c4f3b377bd
>> --- /dev/null
>> +++ b/Documentation/security/landlock/user.rst
>> @@ -0,0 +1,147 @@
>> +================================
>> +Landlock: userland documentation
>> +================================
>> +
>> +Landlock programs
>> +=================
>> +
>> +eBPF programs are used to create security programs.  They are contained and can
>> +call only a whitelist of dedicated functions. Moreover, they can only loop
>> +under strict conditions, which protects from denial of service.  More
>> +information on BPF can be found in *Documentation/networking/filter.txt*.
>> +
>> +
>> +Writing a program
>> +-----------------
>> +
>> +To enforce a security policy, a thread first needs to create a Landlock program.
>> +The easiest way to write an eBPF program depicting a security program is to write
>> +it in the C language.  As described in *samples/bpf/README.rst*, LLVM can
>> +compile such programs.  Files *samples/bpf/landlock1_kern.c* and those in
>> +*tools/testing/selftests/landlock/* can be used as examples.
>> +
>> +Once the eBPF program is created, the next step is to create the metadata
>> +describing the Landlock program.  This metadata includes an expected attach type which
>> +contains the hook type to which the program is tied, and expected attach
>> +triggers which identify the actions for which the program should be run.
>> +
>> +A hook is a policy decision point which exposes the same context type for
>> +each program evaluation.
>> +
>> +A Landlock hook describes the kind of kernel object for which a program will be
>> +triggered to allow or deny an action.  For example, the hook
>> +BPF_LANDLOCK_FS_PICK can be triggered every time a landlocked thread performs a
>> +set of action related to the filesystem (e.g. open, read, write, mount...).
>
>           actions
>
>> +This actions are identified by the `triggers` bitfield.
>> +
>> +The next step is to fill a :c:type:`struct bpf_load_program_attr
>> +<bpf_load_program_attr>` with BPF_PROG_TYPE_LANDLOCK_HOOK, the expected attach
>> +type and other BPF program metadata.  This bpf_attr must then be passed to the
>> +:manpage:`bpf(2)` syscall alongside the BPF_PROG_LOAD command.  If everything
>> +is deemed correct by the kernel, the thread gets a file descriptor referring to
>> +this program.
>> +
>> +In the following code, the *insn* variable is an array of BPF instructions
>> +which can be extracted from an ELF file as is done in bpf_load_file() from
>> +*samples/bpf/bpf_load.c*.
>
> A little confusing.  Is there a mixup of <insn> and <insns>?

Indeed, a typo was inserted with a rewrite of this part.

>
>> +
>> +.. code-block:: c
>> +
>> +    int prog_fd;
>> +    struct bpf_load_program_attr load_attr;
>> +
>> +    memset(&load_attr, 0, sizeof(struct bpf_load_program_attr));
>> +    load_attr.prog_type = BPF_PROG_TYPE_LANDLOCK_HOOK;
>> +    load_attr.expected_attach_type = BPF_LANDLOCK_FS_PICK;
>> +    load_attr.expected_attach_triggers = LANDLOCK_TRIGGER_FS_PICK_OPEN;
>> +    load_attr.insns = insns;
>> +    load_attr.insns_cnt = sizeof(insn) / sizeof(struct bpf_insn);
>> +    load_attr.license = "GPL";
>> +
>> +    prog_fd = bpf_load_program_xattr(&load_attr, log_buf, log_buf_sz);
>> +    if (prog_fd == -1)
>> +        exit(1);
>> +
>> +
>> +Enforcing a program
>> +-------------------
>> +
>> +Once the Landlock program has been created or received (e.g. through a UNIX
>> +socket), the thread willing to sandbox itself (and its future children) should
>> +perform the following two steps.
>> +
>> +The thread should first request to never be allowed to get new privileges with a
>> +call to :manpage:`prctl(2)` and the PR_SET_NO_NEW_PRIVS option.  More
>> +information can be found in *Documentation/prctl/no_new_privs.txt*.
>> +
>> +.. code-block:: c
>> +
>> +    if (prctl(PR_SET_NO_NEW_PRIVS, 1, NULL, 0, 0))
>> +        exit(1);
>> +
>> +A thread can apply a program to itself by using the :manpage:`seccomp(2)` syscall.
>> +The operation is SECCOMP_PREPEND_LANDLOCK_PROG, the flags must be empty and the
>> +*args* argument must point to a valid Landlock program file descriptor.
>> +
>> +.. code-block:: c
>> +
>> +    if (seccomp(SECCOMP_PREPEND_LANDLOCK_PROG, 0, &fd))
>> +        exit(1);
>> +
>> +If the syscall succeeds, the program is now enforced on the calling thread and
>> +will be enforced on all its subsequently created children of the thread as
>> +well.  Once a thread is landlocked, there is no way to remove this security
>> +policy, only stacking more restrictions is allowed.  The program evaluation is
>> +performed from the newest to the oldest.
>> +
>> +When a syscall ask for an action on a kernel object, if this action is denied,
>
>                   asks
>
>> +then an EACCES errno code is returned through the syscall.
>> +
>> +
>> +.. _inherited_programs:
>> +
>> +Inherited programs
>> +------------------
>> +
>> +Every new thread resulting from a :manpage:`clone(2)` inherits Landlock program
>> +restrictions from its parent.  This is similar to the seccomp inheritance as
>> +described in *Documentation/prctl/seccomp_filter.txt*.
>> +
>> +
>> +Ptrace restrictions
>> +-------------------
>> +
>> +A landlocked process has less privileges than a non-landlocked process and must
>> +then be subject to additional restrictions when manipulating another process.
>> +To be allowed to use :manpage:`ptrace(2)` and related syscalls on a target
>> +process, a landlocked process must have a subset of the target process programs.
>             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Maybe that last statement is correct, but it seems to me that it is missing something.

What about this:

To be allowed to trace a process (using :manpage:`ptrace(2)`), a
landlocked tracer process must only be constrained by a subset (possibly
empty) of the Landlock programs which are also applied to the tracee.
This ensure that the tracer has less or the same constraints than the
tracee, hence protecting against privilege escalation.

>
>> +
>> +
>> +Landlock structures and constants
>> +=================================
>> +
>> +Hook types
>> +----------
>> +
>> +.. kernel-doc:: include/uapi/linux/landlock.h
>> +    :functions: landlock_hook_type
>> +
>> +
>> +Contexts
>> +--------
>> +
>> +.. kernel-doc:: include/uapi/linux/landlock.h
>> +    :functions: landlock_ctx_fs_pick landlock_ctx_fs_walk landlock_ctx_fs_get
>> +
>> +
>> +Triggers for fs_pick
>> +--------------------
>> +
>> +.. kernel-doc:: include/uapi/linux/landlock.h
>> +    :functions: landlock_triggers
>> +
>> +
>> +Additional documentation
>> +========================
>> +
>> +See https://landlock.io
>>
>
>

--
Mickaël Salaün
ANSSI/SDE/ST/LAM

Les données à caractère personnel recueillies et traitées dans le cadre de cet échange, le sont à seule fin d’exécution d’une relation professionnelle et s’opèrent dans cette seule finalité et pour la durée nécessaire à cette relation. Si vous souhaitez faire usage de vos droits de consultation, de rectification et de suppression de vos données, veuillez contacter contact.rgpd@sgdsn.gouv.fr. Si vous avez reçu ce message par erreur, nous vous remercions d’en informer l’expéditeur et de détruire le message. The personal data collected and processed during this exchange aims solely at completing a business relationship and is limited to the necessary duration of that relationship. If you wish to use your rights of consultation, rectification and deletion of your data, please contact: contact.rgpd@sgdsn.gouv.fr. If you have received this message in error, we thank you for informing the sender and destroying the message.

^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-07-31 19:11         ` Mickaël Salaün
@ 2019-08-01 17:35           ` Alexei Starovoitov
  2019-08-06 16:24             ` Mickaël Salaün
  0 siblings, 1 reply; 22+ messages in thread
From: Alexei Starovoitov @ 2019-08-01 17:35 UTC (permalink / raw)
  To: Mickaël Salaün
  Cc: Mickaël Salaün, LKML, Alexander Viro,
	Alexei Starovoitov, Andrew Morton, Andy Lutomirski,
	Arnaldo Carvalho de Melo, Casey Schaufler, Daniel Borkmann,
	David Drysdale, David S . Miller, Eric W . Biederman,
	James Morris, Jann Horn, John Johansen, Jonathan Corbet,
	Kees Cook, Michael Kerrisk, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	Kernel Hardening, Linux API, Linux-Fsdevel, LSM List,
	Network Development

On Wed, Jul 31, 2019 at 09:11:10PM +0200, Mickaël Salaün wrote:
> 
> 
> On 31/07/2019 20:58, Alexei Starovoitov wrote:
> > On Wed, Jul 31, 2019 at 11:46 AM Mickaël Salaün
> > <mickael.salaun@ssi.gouv.fr> wrote:
> >>>> +    for (i = 0; i < htab->n_buckets; i++) {
> >>>> +            head = select_bucket(htab, i);
> >>>> +            hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
> >>>> +                    landlock_inode_remove_map(*((struct inode **)l->key), map);
> >>>> +            }
> >>>> +    }
> >>>> +    htab_map_free(map);
> >>>> +}
> >>>
> >>> user space can delete the map.
> >>> that will trigger inode_htab_map_free() which will call
> >>> landlock_inode_remove_map().
> >>> which will simply itereate the list and delete from the list.
> >>
> >> landlock_inode_remove_map() removes the reference to the map (being
> >> freed) from the inode (with an RCU lock).
> >
> > I'm going to ignore everything else for now and focus only on this bit,
> > since it's fundamental issue to address before this discussion can
> > go any further.
> > rcu_lock is not a spin_lock. I'm pretty sure you know this.
> > But you're arguing that it's somehow protecting from the race
> > I mentioned above?
> >
> 
> I was just clarifying your comment to avoid misunderstanding about what
> is being removed.
> 
> As said in the full response, there is currently a race but, if I add a
> bpf_map_inc() call when the map is referenced by inode->security, then I
> don't see how a race could occur because such added map could only be
> freed in a security_inode_free() (as long as it retains a reference to
> this inode).

then it will be a cycle and a map will never be deleted?
closing map_fd should delete a map. It cannot be alive if it's not
pinned in bpffs, there are no FDs that are holding it, and no progs using it.
So the map deletion will iterate over inodes that belong to this map.
In parallel security_inode_free() will be called that will iterate
over its link list that contains elements from different maps.
So the same link list is modified by two cpus.
Where is a lock that protects from concurrent links list manipulations?

> Les données à caractère personnel recueillies et traitées dans le cadre de cet échange, le sont à seule fin d’exécution d’une relation professionnelle et s’opèrent dans cette seule finalité et pour la durée nécessaire à cette relation. Si vous souhaitez faire usage de vos droits de consultation, de rectification et de suppression de vos données, veuillez contacter contact.rgpd@sgdsn.gouv.fr. Si vous avez reçu ce message par erreur, nous vous remercions d’en informer l’expéditeur et de détruire le message. The personal data collected and processed during this exchange aims solely at completing a business relationship and is limited to the necessary duration of that relationship. If you wish to use your rights of consultation, rectification and deletion of your data, please contact: contact.rgpd@sgdsn.gouv.fr. If you have received this message in error, we thank you for informing the sender and destroying the message.

Please get rid of this. It's absolutely not appropriate on public mailing list.
Next time I'd have to ignore emails that contain such disclaimers.


^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation for Landlock
  2019-08-01 17:03     ` Mickaël Salaün
@ 2019-08-01 17:49       ` Randy Dunlap
  0 siblings, 0 replies; 22+ messages in thread
From: Randy Dunlap @ 2019-08-01 17:49 UTC (permalink / raw)
  To: Mickaël Salaün, Mickaël Salaün, linux-kernel
  Cc: Alexander Viro, Alexei Starovoitov, Andrew Morton,
	Andy Lutomirski, Arnaldo Carvalho de Melo, Casey Schaufler,
	Daniel Borkmann, David Drysdale, David S . Miller,
	Eric W . Biederman, James Morris, Jann Horn, John Johansen,
	Jonathan Corbet, Kees Cook, Michael Kerrisk, Paul Moore,
	Sargun Dhillon, Serge E . Hallyn, Shuah Khan, Stephen Smalley,
	Tejun Heo, Tetsuo Handa, Thomas Graf, Tycho Andersen,
	Will Drewry, kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

On 8/1/19 10:03 AM, Mickaël Salaün wrote:
>>> +Ptrace restrictions
>>> +-------------------
>>> +
>>> +A landlocked process has less privileges than a non-landlocked process and must
>>> +then be subject to additional restrictions when manipulating another process.
>>> +To be allowed to use :manpage:`ptrace(2)` and related syscalls on a target
>>> +process, a landlocked process must have a subset of the target process programs.
>>             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> Maybe that last statement is correct, but it seems to me that it is missing something.
> What about this:
> 
> To be allowed to trace a process (using :manpage:`ptrace(2)`), a
> landlocked tracer process must only be constrained by a subset (possibly
> empty) of the Landlock programs which are also applied to the tracee.
> This ensure that the tracer has less or the same constraints than the

       ensures

> tracee, hence protecting against privilege escalation.

Yes, better.  Thanks.


-- 
~Randy

^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-08-01 17:35           ` Alexei Starovoitov
@ 2019-08-06 16:24             ` Mickaël Salaün
  0 siblings, 0 replies; 22+ messages in thread
From: Mickaël Salaün @ 2019-08-06 16:24 UTC (permalink / raw)
  To: Alexei Starovoitov
  Cc: Mickaël Salaün, LKML, Alexander Viro,
	Alexei Starovoitov, Andrew Morton, Andy Lutomirski,
	Arnaldo Carvalho de Melo, Casey Schaufler, Daniel Borkmann,
	David Drysdale, David S . Miller, Eric W . Biederman,
	James Morris, Jann Horn, John Johansen, Jonathan Corbet,
	Kees Cook, Michael Kerrisk, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	Kernel Hardening, Linux API, Linux-Fsdevel, LSM List,
	Network Development


On 01/08/2019 19:35, Alexei Starovoitov wrote:
> On Wed, Jul 31, 2019 at 09:11:10PM +0200, Mickaël Salaün wrote:
>>
>>
>> On 31/07/2019 20:58, Alexei Starovoitov wrote:
>>> On Wed, Jul 31, 2019 at 11:46 AM Mickaël Salaün
>>> <mickael.salaun@ssi.gouv.fr> wrote:
>>>>>> +    for (i = 0; i < htab->n_buckets; i++) {
>>>>>> +            head = select_bucket(htab, i);
>>>>>> +            hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
>>>>>> +                    landlock_inode_remove_map(*((struct inode **)l->key), map);
>>>>>> +            }
>>>>>> +    }
>>>>>> +    htab_map_free(map);
>>>>>> +}
>>>>>
>>>>> user space can delete the map.
>>>>> that will trigger inode_htab_map_free() which will call
>>>>> landlock_inode_remove_map().
>>>>> which will simply itereate the list and delete from the list.
>>>>
>>>> landlock_inode_remove_map() removes the reference to the map (being
>>>> freed) from the inode (with an RCU lock).
>>>
>>> I'm going to ignore everything else for now and focus only on this bit,
>>> since it's fundamental issue to address before this discussion can
>>> go any further.
>>> rcu_lock is not a spin_lock. I'm pretty sure you know this.
>>> But you're arguing that it's somehow protecting from the race
>>> I mentioned above?
>>>
>>
>> I was just clarifying your comment to avoid misunderstanding about what
>> is being removed.
>>
>> As said in the full response, there is currently a race but, if I add a
>> bpf_map_inc() call when the map is referenced by inode->security, then I
>> don't see how a race could occur because such added map could only be
>> freed in a security_inode_free() (as long as it retains a reference to
>> this inode).
> 
> then it will be a cycle and a map will never be deleted?
> closing map_fd should delete a map. It cannot be alive if it's not
> pinned in bpffs, there are no FDs that are holding it, and no progs using it.
> So the map deletion will iterate over inodes that belong to this map.
> In parallel security_inode_free() will be called that will iterate
> over its link list that contains elements from different maps.
> So the same link list is modified by two cpus.
> Where is a lock that protects from concurrent links list manipulations?

Ok, I think I got it. What about this fix?

diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
index 4fc7755042f0..3226e50b6211 100644
--- a/kernel/bpf/hashtab.c
+++ b/kernel/bpf/hashtab.c
@@ -1708,10 +1708,16 @@ static void inode_htab_map_free(struct bpf_map *map)

 	for (i = 0; i < htab->n_buckets; i++) {
 		head = select_bucket(htab, i);
-		hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
+		rcu_read_lock();
+		hlist_nulls_for_each_entry_rcu(l, n, head, hash_node) {
 			landlock_inode_remove_map(*((struct inode **)l->key), map);
 		}
+		rcu_read_unlock();
 	}
+	/*
+	 * The last pending put_landlock_inode_map() may be called here, before
+	 * the rcu_barrier() from htab_map_free().
+	 */
 	htab_map_free(map);
 }

diff --git a/security/landlock/common.h b/security/landlock/common.h
index b0ba3f31ac7d..535c6a4292b9 100644
--- a/security/landlock/common.h
+++ b/security/landlock/common.h
@@ -58,6 +58,11 @@ struct landlock_prog_set {
 	refcount_t usage;
 };

+struct landlock_inode_security {
+	struct list_head list;
+	spinlock_t lock;
+};
+
 struct landlock_inode_map {
 	struct list_head list;
 	struct rcu_head rcu_put;
diff --git a/security/landlock/hooks_fs.c b/security/landlock/hooks_fs.c
index 8c9d6a333111..b9bfd558f8b8 100644
--- a/security/landlock/hooks_fs.c
+++ b/security/landlock/hooks_fs.c
@@ -10,6 +10,7 @@
 #include <linux/kernel.h> /* ARRAY_SIZE */
 #include <linux/lsm_hooks.h>
 #include <linux/rcupdate.h> /* synchronize_rcu() */
+#include <linux/spinlock.h>
 #include <linux/stat.h> /* S_ISDIR */
 #include <linux/stddef.h> /* offsetof */
 #include <linux/types.h> /* uintptr_t */
@@ -251,13 +252,16 @@ static int hook_sb_pivotroot(const struct path *old_path,

 /* inode helpers */

-static inline struct list_head *inode_landlock(const struct inode *inode)
+static inline struct landlock_inode_security *inode_landlock(
+		const struct inode *inode)
 {
 	return inode->i_security + landlock_blob_sizes.lbs_inode;
 }

 int landlock_inode_add_map(struct inode *inode, struct bpf_map *map)
 {
+	unsigned long flags;
+	struct landlock_inode_security *inode_sec = inode_landlock(inode);
 	struct landlock_inode_map *inode_map;

 	inode_map = kzalloc(sizeof(*inode_map), GFP_ATOMIC);
@@ -266,60 +270,66 @@ int landlock_inode_add_map(struct inode *inode, struct bpf_map *map)
 	INIT_LIST_HEAD(&inode_map->list);
 	inode_map->map = map;
 	inode_map->inode = inode;
-	list_add_tail(&inode_map->list, inode_landlock(inode));
+	spin_lock_irqsave(&inode_sec->lock, flags);
+	list_add_tail_rcu(&inode_map->list, &inode_sec->list);
+	spin_unlock_irqrestore(&inode_sec->lock, flags);
 	return 0;
 }

 static void put_landlock_inode_map(struct rcu_head *head)
 {
 	struct landlock_inode_map *inode_map;
-	int err;

 	inode_map = container_of(head, struct landlock_inode_map, rcu_put);
-	err = bpf_inode_ptr_unlocked_htab_map_delete_elem(inode_map->map,
+	bpf_inode_ptr_unlocked_htab_map_delete_elem(inode_map->map,
 			&inode_map->inode, false);
-	bpf_map_put(inode_map->map);
 	kfree(inode_map);
 }

 void landlock_inode_remove_map(struct inode *inode, const struct bpf_map *map)
 {
+	unsigned long flags;
+	struct landlock_inode_security *inode_sec = inode_landlock(inode);
 	struct landlock_inode_map *inode_map;
-	bool found = false;

+	spin_lock_irqsave(&inode_sec->lock, flags);
 	rcu_read_lock();
-	list_for_each_entry_rcu(inode_map, inode_landlock(inode), list) {
+	list_for_each_entry_rcu(inode_map, &inode_sec->list, list) {
 		if (inode_map->map == map) {
-			found = true;
 			list_del_rcu(&inode_map->list);
 			kfree_rcu(inode_map, rcu_put);
 			break;
 		}
 	}
 	rcu_read_unlock();
-	WARN_ON(!found);
+	spin_unlock_irqrestore(&inode_sec->lock, flags);
 }

 /* inode hooks */

 static int hook_inode_alloc_security(struct inode *inode)
 {
-	struct list_head *ll_inode = inode_landlock(inode);
+	struct landlock_inode_security *inode_sec = inode_landlock(inode);

-	INIT_LIST_HEAD(ll_inode);
+	INIT_LIST_HEAD(&inode_sec->list);
+	spin_lock_init(&inode_sec->lock);
 	return 0;
 }

 static void hook_inode_free_security(struct inode *inode)
 {
+	unsigned long flags;
+	struct landlock_inode_security *inode_sec = inode_landlock(inode);
 	struct landlock_inode_map *inode_map;

+	spin_lock_irqsave(&inode_sec->lock, flags);
 	rcu_read_lock();
-	list_for_each_entry_rcu(inode_map, inode_landlock(inode), list) {
+	list_for_each_entry_rcu(inode_map, &inode_sec->list, list) {
 		list_del_rcu(&inode_map->list);
 		call_rcu(&inode_map->rcu_put, put_landlock_inode_map);
 	}
 	rcu_read_unlock();
+	spin_unlock_irqrestore(&inode_sec->lock, flags);
 }

 /* a directory inode contains only one dentry */
diff --git a/security/landlock/init.c b/security/landlock/init.c
index 35165fc8a595..1305255f5d2e 100644
--- a/security/landlock/init.c
+++ b/security/landlock/init.c
@@ -137,7 +137,7 @@ static int __init landlock_init(void)
 }

 struct lsm_blob_sizes landlock_blob_sizes __lsm_ro_after_init = {
-	.lbs_inode = sizeof(struct list_head),
+	.lbs_inode = sizeof(struct landlock_inode_security),
 };

 DEFINE_LSM(LANDLOCK_NAME) = {


> 
>> Les données à caractère personnel recueillies et traitées dans le cadre de cet échange, le sont à seule fin d’exécution d’une relation professionnelle et s’opèrent dans cette seule finalité et pour la durée nécessaire à cette relation. Si vous souhaitez faire usage de vos droits de consultation, de rectification et de suppression de vos données, veuillez contacter contact.rgpd@sgdsn.gouv.fr. Si vous avez reçu ce message par erreur, nous vous remercions d’en informer l’expéditeur et de détruire le message. The personal data collected and processed during this exchange aims solely at completing a business relationship and is limited to the necessary duration of that relationship. If you wish to use your rights of consultation, rectification and deletion of your data, please contact: contact.rgpd@sgdsn.gouv.fr. If you have received this message in error, we thank you for informing the sender and destroying the message.
> 
> Please get rid of this. It's absolutely not appropriate on public mailing list.
> Next time I'd have to ignore emails that contain such disclaimers.

Unfortunately this message is automatically appended (server-side) to all my
professional emails...

^ permalink raw reply related	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-07-31 18:46     ` Mickaël Salaün
  2019-07-31 18:58       ` Alexei Starovoitov
@ 2019-09-08 22:09       ` Mickaël Salaün
  2019-09-08 22:19         ` Al Viro
  1 sibling, 1 reply; 22+ messages in thread
From: Mickaël Salaün @ 2019-09-08 22:09 UTC (permalink / raw)
  To: Alexei Starovoitov
  Cc: Mickaël Salaün, linux-kernel, Alexander Viro,
	Alexei Starovoitov, Andrew Morton, Andy Lutomirski,
	Arnaldo Carvalho de Melo, Casey Schaufler, Daniel Borkmann,
	David Drysdale, David S . Miller, Eric W . Biederman,
	James Morris, Jann Horn, John Johansen, Jonathan Corbet,
	Kees Cook, Michael Kerrisk, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev



On 31/07/2019 20:46, Mickaël Salaün wrote:
> 
> On 27/07/2019 03:40, Alexei Starovoitov wrote:
>> On Sun, Jul 21, 2019 at 11:31:12PM +0200, Mickaël Salaün wrote:
>>> FIXME: 64-bits in the doc
> 
> FYI, this FIXME was fixed, just not removed from this message. :)
> 
>>>
>>> This new map store arbitrary values referenced by inode keys.  The map
>>> can be updated from user space with file descriptor pointing to inodes
>>> tied to a file system.  From an eBPF (Landlock) program point of view,
>>> such a map is read-only and can only be used to retrieved a value tied
>>> to a given inode.  This is useful to recognize an inode tagged by user
>>> space, without access right to this inode (i.e. no need to have a write
>>> access to this inode).
>>>
>>> Add dedicated BPF functions to handle this type of map:
>>> * bpf_inode_htab_map_update_elem()
>>> * bpf_inode_htab_map_lookup_elem()
>>> * bpf_inode_htab_map_delete_elem()
>>>
>>> This new map require a dedicated helper inode_map_lookup_elem() because
>>> of the key which is a pointer to an opaque data (only provided by the
>>> kernel).  This act like a (physical or cryptographic) key, which is why
>>> it is also not allowed to get the next key.
>>>
>>> Signed-off-by: Mickaël Salaün <mic@digikod.net>
>>
>> there are too many things to comment on.
>> Let's do this patch.
>>
>> imo inode_map concept is interesting, but see below...
>>
>>> +
>>> +    /*
>>> +     * Limit number of entries in an inode map to the maximum number of
>>> +     * open files for the current process. The maximum number of file
>>> +     * references (including all inode maps) for a process is then
>>> +     * (RLIMIT_NOFILE - 1) * RLIMIT_NOFILE. If the process' RLIMIT_NOFILE
>>> +     * is 0, then any entry update is forbidden.
>>> +     *
>>> +     * An eBPF program can inherit all the inode map FD. The worse case is
>>> +     * to fill a bunch of arraymaps, create an eBPF program, close the
>>> +     * inode map FDs, and start again. The maximum number of inode map
>>> +     * entries can then be close to RLIMIT_NOFILE^3.
>>> +     */
>>> +    if (attr->max_entries > rlimit(RLIMIT_NOFILE))
>>> +            return -EMFILE;
>>
>> rlimit is checked, but no fd are consumed.
>> Once created such inode map_fd can be passed to a different process.
>> map_fd can be pinned into bpffs.
>> etc.
>> what the value of the check?
> 
> I was looking for the most meaningful limit for a process, and rlimit is
> the best I found. As the limit of open FD per processes, rlimit is not
> perfect, but I think the semantic is close here (e.g. a process can also
> pass FD through unix socket).
> 
>>
>>> +
>>> +    /* decorelate UAPI from kernel API */
>>> +    attr->key_size = sizeof(struct inode *);
>>> +
>>> +    return htab_map_alloc_check(attr);
>>> +}
>>> +
>>> +static void inode_htab_put_key(void *key)
>>> +{
>>> +    struct inode **inode = key;
>>> +
>>> +    if ((*inode)->i_state & I_FREEING)
>>> +            return;
>>
>> checking the state without take a lock? isn't it racy?
> 
> This should only trigger when called from security_inode_free(). I'll
> add a comment.
> 
>>
>>> +    iput(*inode);
>>> +}
>>> +
>>> +/* called from syscall or (never) from eBPF program */
>>> +static int map_get_next_no_key(struct bpf_map *map, void *key, void *next_key)
>>> +{
>>> +    /* do not leak a file descriptor */
>>
>> what this comment suppose to mean?
> 
> Because a key is a reference to an inode, a possible return value for
> this function could be a file descriptor pointing to this inode (the
> same way a file descriptor is use to add an element). For now, I don't
> want to implement a way for a process with such a map to extract such
> inode, which I compare to a possible leak (of information, not kernel
> memory nor object). This could be implemented in the future if there is
> value in it (and probably some additional safeguards), though.
> 
>>
>>> +    return -ENOTSUPP;
>>> +}
>>> +
>>> +/* must call iput(inode) after this call */
>>> +static struct inode *inode_from_fd(int ufd, bool check_access)
>>> +{
>>> +    struct inode *ret;
>>> +    struct fd f;
>>> +    int deny;
>>> +
>>> +    f = fdget(ufd);
>>> +    if (unlikely(!f.file))
>>> +            return ERR_PTR(-EBADF);
>>> +    /* TODO?: add this check when called from an eBPF program too (already
>>> +    * checked by the LSM parent hooks anyway) */
>>> +    if (unlikely(IS_PRIVATE(file_inode(f.file)))) {
>>> +            ret = ERR_PTR(-EINVAL);
>>> +            goto put_fd;
>>> +    }
>>> +    /* check if the FD is tied to a mount point */
>>> +    /* TODO?: add this check when called from an eBPF program too */
>>> +    if (unlikely(f.file->f_path.mnt->mnt_flags & MNT_INTERNAL)) {
>>> +            ret = ERR_PTR(-EINVAL);
>>> +            goto put_fd;
>>> +    }
>>
>> a bunch of TODOs do not inspire confidence.
> 
> I think the current implement is good, but these TODOs are here to draw
> attention on particular points for which I would like external review
> and opinion (hence the "?").
> 
>>
>>> +    if (check_access) {
>>> +            /*
>>> +            * must be allowed to access attributes from this file to then
>>> +            * be able to compare an inode to its map entry
>>> +            */
>>> +            deny = security_inode_getattr(&f.file->f_path);
>>> +            if (deny) {
>>> +                    ret = ERR_PTR(deny);
>>> +                    goto put_fd;
>>> +            }
>>> +    }
>>> +    ret = file_inode(f.file);
>>> +    ihold(ret);
>>> +
>>> +put_fd:
>>> +    fdput(f);
>>> +    return ret;
>>> +}
>>> +
>>> +/*
>>> + * The key is a FD when called from a syscall, but an inode address when called
>>> + * from an eBPF program.
>>> + */
>>> +
>>> +/* called from syscall */
>>> +int bpf_inode_fd_htab_map_lookup_elem(struct bpf_map *map, int *key, void *value)
>>> +{
>>> +    void *ptr;
>>> +    struct inode *inode;
>>> +    int ret;
>>> +
>>> +    /* check inode access */
>>> +    inode = inode_from_fd(*key, true);
>>> +    if (IS_ERR(inode))
>>> +            return PTR_ERR(inode);
>>> +
>>> +    rcu_read_lock();
>>> +    ptr = htab_map_lookup_elem(map, &inode);
>>> +    iput(inode);
>>> +    if (IS_ERR(ptr)) {
>>> +            ret = PTR_ERR(ptr);
>>> +    } else if (!ptr) {
>>> +            ret = -ENOENT;
>>> +    } else {
>>> +            ret = 0;
>>> +            copy_map_value(map, value, ptr);
>>> +    }
>>> +    rcu_read_unlock();
>>> +    return ret;
>>> +}
>>> +
>>> +/* called from kernel */
>>
>> wrong comment?
>> kernel side cannot call it, right?
> 
> This is called from bpf_inode_fd_htab_map_delete_elem() (code just
> beneath), and from
> kernel/bpf/syscall.c:bpf_inode_ptr_unlocked_htab_map_delet_elem() which
> can be called by security_inode_free() (hook_inode_free_security).
> 
>>
>>> +int bpf_inode_ptr_locked_htab_map_delete_elem(struct bpf_map *map,
>>> +            struct inode **key, bool remove_in_inode)
>>> +{
>>> +    if (remove_in_inode)
>>> +            landlock_inode_remove_map(*key, map);
>>> +    return htab_map_delete_elem(map, key);
>>> +}
>>> +
>>> +/* called from syscall */
>>> +int bpf_inode_fd_htab_map_delete_elem(struct bpf_map *map, int *key)
>>> +{
>>> +    struct inode *inode;
>>> +    int ret;
>>> +
>>> +    /* do not check inode access (similar to directory check) */
>>> +    inode = inode_from_fd(*key, false);
>>> +    if (IS_ERR(inode))
>>> +            return PTR_ERR(inode);
>>> +    ret = bpf_inode_ptr_locked_htab_map_delete_elem(map, &inode, true);
>>> +    iput(inode);
>>> +    return ret;
>>> +}
>>> +
>>> +/* called from syscall */
>>> +int bpf_inode_fd_htab_map_update_elem(struct bpf_map *map, int *key, void *value,
>>> +            u64 map_flags)
>>> +{
>>> +    struct inode *inode;
>>> +    int ret;
>>> +
>>> +    WARN_ON_ONCE(!rcu_read_lock_held());
>>> +
>>> +    /* check inode access */
>>> +    inode = inode_from_fd(*key, true);
>>> +    if (IS_ERR(inode))
>>> +            return PTR_ERR(inode);
>>> +    ret = htab_map_update_elem(map, &inode, value, map_flags);
>>> +    if (!ret)
>>> +            ret = landlock_inode_add_map(inode, map);
>>> +    iput(inode);
>>> +    return ret;
>>> +}
>>> +
>>> +static void inode_htab_map_free(struct bpf_map *map)
>>> +{
>>> +    struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
>>> +    struct hlist_nulls_node *n;
>>> +    struct hlist_nulls_head *head;
>>> +    struct htab_elem *l;
>>> +    int i;
>>> +
>>> +    for (i = 0; i < htab->n_buckets; i++) {
>>> +            head = select_bucket(htab, i);
>>> +            hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
>>> +                    landlock_inode_remove_map(*((struct inode **)l->key), map);
>>> +            }
>>> +    }
>>> +    htab_map_free(map);
>>> +}
>>
>> user space can delete the map.
>> that will trigger inode_htab_map_free() which will call
>> landlock_inode_remove_map().
>> which will simply itereate the list and delete from the list.
> 
> landlock_inode_remove_map() removes the reference to the map (being
> freed) from the inode (with an RCU lock).
> 
>>
>> While in parallel inode can be destoyed and hook_inode_free_security()
>> will be called.
>> I think nothing that protects from this race.
> 
> According to security_inode_free(), the inode is effectively freed after
> the RCU grace period. However, I forgot to call bpf_map_inc() in
> landlock_inode_add_map(), which would prevent the map to be freed
> outside of the security_inode_free(). I'll fix that.
> 
>>
>>> +
>>> +/*
>>> + * We need a dedicated helper to deal with inode maps because the key is a
>>> + * pointer to an opaque data, only provided by the kernel.  This really act
>>> + * like a (physical or cryptographic) key, which is why it is also not allowed
>>> + * to get the next key with map_get_next_key().
>>
>> inode pointer is like cryptographic key? :)
> 
> I wanted to highlight the fact that, contrary to other map key types,
> the value of this one should not be readable, only usable. A "secret
> value" is more appropriate but still confusing. I'll rephrase that.
> 
>>
>>> + */
>>> +BPF_CALL_2(bpf_inode_map_lookup_elem, struct bpf_map *, map, void *, key)
>>> +{
>>> +    WARN_ON_ONCE(!rcu_read_lock_held());
>>> +    return (unsigned long)htab_map_lookup_elem(map, &key);
>>> +}
>>> +
>>> +const struct bpf_func_proto bpf_inode_map_lookup_elem_proto = {
>>> +    .func           = bpf_inode_map_lookup_elem,
>>> +    .gpl_only       = false,
>>> +    .pkt_access     = true,
>>
>> pkt_access ? :)
> 
> This slipped in with this rebase, I'll remove it. :)
> 
>>
>>> +    .ret_type       = RET_PTR_TO_MAP_VALUE_OR_NULL,
>>> +    .arg1_type      = ARG_CONST_MAP_PTR,
>>> +    .arg2_type      = ARG_PTR_TO_INODE,
>>> +};
>>> diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
>>> index b2a8cb14f28e..e46441c42b68 100644
>>> --- a/kernel/bpf/syscall.c
>>> +++ b/kernel/bpf/syscall.c
>>> @@ -801,6 +801,8 @@ static int map_lookup_elem(union bpf_attr *attr)
>>>      } else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
>>>                 map->map_type == BPF_MAP_TYPE_STACK) {
>>>              err = map->ops->map_peek_elem(map, value);
>>> +    } else if (map->map_type == BPF_MAP_TYPE_INODE) {
>>> +            err = bpf_inode_fd_htab_map_lookup_elem(map, key, value);
>>>      } else {
>>>              rcu_read_lock();
>>>              if (map->ops->map_lookup_elem_sys_only)
>>> @@ -951,6 +953,10 @@ static int map_update_elem(union bpf_attr *attr)
>>>      } else if (map->map_type == BPF_MAP_TYPE_QUEUE ||
>>>                 map->map_type == BPF_MAP_TYPE_STACK) {
>>>              err = map->ops->map_push_elem(map, value, attr->flags);
>>> +    } else if (map->map_type == BPF_MAP_TYPE_INODE) {
>>> +            rcu_read_lock();
>>> +            err = bpf_inode_fd_htab_map_update_elem(map, key, value, attr->flags);
>>> +            rcu_read_unlock();
>>>      } else {
>>>              rcu_read_lock();
>>>              err = map->ops->map_update_elem(map, key, value, attr->flags);
>>> @@ -1006,7 +1012,10 @@ static int map_delete_elem(union bpf_attr *attr)
>>>      preempt_disable();
>>>      __this_cpu_inc(bpf_prog_active);
>>>      rcu_read_lock();
>>> -    err = map->ops->map_delete_elem(map, key);
>>> +    if (map->map_type == BPF_MAP_TYPE_INODE)
>>> +            err = bpf_inode_fd_htab_map_delete_elem(map, key);
>>> +    else
>>> +            err = map->ops->map_delete_elem(map, key);
>>>      rcu_read_unlock();
>>>      __this_cpu_dec(bpf_prog_active);
>>>      preempt_enable();
>>> @@ -1018,6 +1027,22 @@ static int map_delete_elem(union bpf_attr *attr)
>>>      return err;
>>>  }
>>>
>>> +int bpf_inode_ptr_unlocked_htab_map_delete_elem(struct bpf_map *map,
>>> +                                            struct inode **key, bool remove_in_inode)
>>> +{
>>> +    int err;
>>> +
>>> +    preempt_disable();
>>> +    __this_cpu_inc(bpf_prog_active);
>>> +    rcu_read_lock();
>>> +    err = bpf_inode_ptr_locked_htab_map_delete_elem(map, key, remove_in_inode);
>>> +    rcu_read_unlock();
>>> +    __this_cpu_dec(bpf_prog_active);
>>> +    preempt_enable();
>>> +    maybe_wait_bpf_programs(map);
>>
>> if that function was actually doing synchronize_rcu() the consequences
>> would have been unpleasant. Fortunately it's a nop in this case.
>> Please read the code carefully before copy-paste.
>> Also what do you think the reason of bpf_prog_active above?
>> What is the reason of rcu_read_lock above?
> 
> The RCU is used as for every map modifications (usually from userspace).
> I wasn't sure about the other protections so I kept the same (generic)
> checks as in map_delete_elem() (just above) because this function follow
> the same semantic. What can I safely remove?
> 
>>
>> I think the patch set needs to shrink at least in half to be reviewable.
>> The way you tie seccomp and lsm is probably the biggest obstacle
>> than any of the bugs above.
>> Can you drop seccomp ? and do it as normal lsm ?
> 
> The seccomp/enforcement part is needed to have a minimum viable product,
> i.e. a process able to sandbox itself. Are you suggesting to first merge
> a version when it is only possible to create inode maps but not use them
> in an useful way (i.e. for sandboxing)? I can do it if it's OK with you,
> and I hope it will not be a problem for the security folks if it can
> help to move forward.

I talked with Kees Cook and James Morris at LSS NA, and I think the
better strategy to shrink this patch series is to tackle a much less
complex problem at first. Instead on focusing right now on file system,
the next version of this patch series will focus on memory protection,
which is also something desired. I'll then iterate with file system
support (i.e. inode maps) and other use cases once the basics of
Landlock are upstream. For this next series, the majority of the code
will be on the LSM side, while the eBPF part will mainly consist to add
a new program type. Because bpf-next is moving rapidly, I think it still
make sense to base this work on this tree (instead of linux-security).

Regards,
 Mickaël

^ permalink raw reply	[flat|nested] 22+ messages in thread

* Re: [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode
  2019-09-08 22:09       ` Mickaël Salaün
@ 2019-09-08 22:19         ` Al Viro
  0 siblings, 0 replies; 22+ messages in thread
From: Al Viro @ 2019-09-08 22:19 UTC (permalink / raw)
  To: Mickaël Salaün
  Cc: Alexei Starovoitov, Mickaël Salaün, linux-kernel,
	Alexei Starovoitov, Andrew Morton, Andy Lutomirski,
	Arnaldo Carvalho de Melo, Casey Schaufler, Daniel Borkmann,
	David Drysdale, David S . Miller, Eric W . Biederman,
	James Morris, Jann Horn, John Johansen, Jonathan Corbet,
	Kees Cook, Michael Kerrisk, Paul Moore, Sargun Dhillon,
	Serge E . Hallyn, Shuah Khan, Stephen Smalley, Tejun Heo,
	Tetsuo Handa, Thomas Graf, Tycho Andersen, Will Drewry,
	kernel-hardening, linux-api, linux-fsdevel,
	linux-security-module, netdev

On Mon, Sep 09, 2019 at 12:09:57AM +0200, Mickaël Salaün wrote:

> >>> +    rcu_read_lock();
> >>> +    ptr = htab_map_lookup_elem(map, &inode);
> >>> +    iput(inode);

Wait a sec.  You are doing _what_ under rcu_read_lock()?

> >>> +    if (IS_ERR(ptr)) {
> >>> +            ret = PTR_ERR(ptr);
> >>> +    } else if (!ptr) {
> >>> +            ret = -ENOENT;
> >>> +    } else {
> >>> +            ret = 0;
> >>> +            copy_map_value(map, value, ptr);
> >>> +    }
> >>> +    rcu_read_unlock();

^ permalink raw reply	[flat|nested] 22+ messages in thread

end of thread, other threads:[~2019-09-08 22:20 UTC | newest]

Thread overview: 22+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-07-21 21:31 [PATCH bpf-next v10 00/10] Landlock LSM: Toward unprivileged sandboxing Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 01/10] fs,security: Add a new file access type: MAY_CHROOT Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 02/10] bpf: Add expected_attach_triggers and a is_valid_triggers() verifier Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 03/10] bpf,landlock: Define an eBPF program type for Landlock hooks Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 04/10] seccomp,landlock: Enforce Landlock programs per process hierarchy Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 05/10] landlock: Handle filesystem access control Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 06/10] bpf,landlock: Add a new map type: inode Mickaël Salaün
2019-07-27  1:40   ` Alexei Starovoitov
2019-07-31 18:46     ` Mickaël Salaün
2019-07-31 18:58       ` Alexei Starovoitov
2019-07-31 19:11         ` Mickaël Salaün
2019-08-01 17:35           ` Alexei Starovoitov
2019-08-06 16:24             ` Mickaël Salaün
2019-09-08 22:09       ` Mickaël Salaün
2019-09-08 22:19         ` Al Viro
2019-07-21 21:31 ` [PATCH bpf-next v10 07/10] landlock: Add ptrace restrictions Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 08/10] bpf: Add a Landlock sandbox example Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 09/10] bpf,landlock: Add tests for Landlock Mickaël Salaün
2019-07-21 21:31 ` [PATCH bpf-next v10 10/10] landlock: Add user and kernel documentation " Mickaël Salaün
2019-07-31  1:53   ` Randy Dunlap
2019-08-01 17:03     ` Mickaël Salaün
2019-08-01 17:49       ` Randy Dunlap

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).