Commit c061f33f authored by Kees Cook's avatar Kees Cook Committed by Jonathan Corbet

doc: ReSTify seccomp_filter.txt

This updates seccomp_filter.txt for ReST markup, and moves it under the
user-space API index, since it describes how application author can use
seccomp.
Signed-off-by: default avatarKees Cook <keescook@chromium.org>
Signed-off-by: default avatarJonathan Corbet <corbet@lwn.net>
parent 5e33994d
...@@ -16,6 +16,7 @@ place where this information is gathered. ...@@ -16,6 +16,7 @@ place where this information is gathered.
.. toctree:: .. toctree::
:maxdepth: 2 :maxdepth: 2
seccomp_filter
unshare unshare
.. only:: subproject and html .. only:: subproject and html
......
SECure COMPuting with filters ===========================================
============================= Seccomp BPF (SECure COMPuting with filters)
===========================================
Introduction Introduction
------------ ============
A large number of system calls are exposed to every userland process A large number of system calls are exposed to every userland process
with many of them going unused for the entire lifetime of the process. with many of them going unused for the entire lifetime of the process.
...@@ -27,7 +28,7 @@ pointers which constrains all filters to solely evaluating the system ...@@ -27,7 +28,7 @@ pointers which constrains all filters to solely evaluating the system
call arguments directly. call arguments directly.
What it isn't What it isn't
------------- =============
System call filtering isn't a sandbox. It provides a clearly defined System call filtering isn't a sandbox. It provides a clearly defined
mechanism for minimizing the exposed kernel surface. It is meant to be mechanism for minimizing the exposed kernel surface. It is meant to be
...@@ -40,13 +41,13 @@ system calls in socketcall() is allowed, for instance) which could be ...@@ -40,13 +41,13 @@ system calls in socketcall() is allowed, for instance) which could be
construed, incorrectly, as a more complete sandboxing solution. construed, incorrectly, as a more complete sandboxing solution.
Usage Usage
----- =====
An additional seccomp mode is added and is enabled using the same An additional seccomp mode is added and is enabled using the same
prctl(2) call as the strict seccomp. If the architecture has prctl(2) call as the strict seccomp. If the architecture has
CONFIG_HAVE_ARCH_SECCOMP_FILTER, then filters may be added as below: ``CONFIG_HAVE_ARCH_SECCOMP_FILTER``, then filters may be added as below:
PR_SET_SECCOMP: ``PR_SET_SECCOMP``:
Now takes an additional argument which specifies a new filter Now takes an additional argument which specifies a new filter
using a BPF program. using a BPF program.
The BPF program will be executed over struct seccomp_data The BPF program will be executed over struct seccomp_data
...@@ -55,24 +56,25 @@ PR_SET_SECCOMP: ...@@ -55,24 +56,25 @@ PR_SET_SECCOMP:
acceptable values to inform the kernel which action should be acceptable values to inform the kernel which action should be
taken. taken.
Usage: Usage::
prctl(PR_SET_SECCOMP, SECCOMP_MODE_FILTER, prog); prctl(PR_SET_SECCOMP, SECCOMP_MODE_FILTER, prog);
The 'prog' argument is a pointer to a struct sock_fprog which The 'prog' argument is a pointer to a struct sock_fprog which
will contain the filter program. If the program is invalid, the will contain the filter program. If the program is invalid, the
call will return -1 and set errno to EINVAL. call will return -1 and set errno to ``EINVAL``.
If fork/clone and execve are allowed by @prog, any child If ``fork``/``clone`` and ``execve`` are allowed by @prog, any child
processes will be constrained to the same filters and system processes will be constrained to the same filters and system
call ABI as the parent. call ABI as the parent.
Prior to use, the task must call prctl(PR_SET_NO_NEW_PRIVS, 1) or Prior to use, the task must call ``prctl(PR_SET_NO_NEW_PRIVS, 1)`` or
run with CAP_SYS_ADMIN privileges in its namespace. If these are not run with ``CAP_SYS_ADMIN`` privileges in its namespace. If these are not
true, -EACCES will be returned. This requirement ensures that filter true, ``-EACCES`` will be returned. This requirement ensures that filter
programs cannot be applied to child processes with greater privileges programs cannot be applied to child processes with greater privileges
than the task that installed them. than the task that installed them.
Additionally, if prctl(2) is allowed by the attached filter, Additionally, if ``prctl(2)`` is allowed by the attached filter,
additional filters may be layered on which will increase evaluation additional filters may be layered on which will increase evaluation
time, but allow for further decreasing the attack surface during time, but allow for further decreasing the attack surface during
execution of a process. execution of a process.
...@@ -80,51 +82,52 @@ PR_SET_SECCOMP: ...@@ -80,51 +82,52 @@ PR_SET_SECCOMP:
The above call returns 0 on success and non-zero on error. The above call returns 0 on success and non-zero on error.
Return values Return values
------------- =============
A seccomp filter may return any of the following values. If multiple A seccomp filter may return any of the following values. If multiple
filters exist, the return value for the evaluation of a given system filters exist, the return value for the evaluation of a given system
call will always use the highest precedent value. (For example, call will always use the highest precedent value. (For example,
SECCOMP_RET_KILL will always take precedence.) ``SECCOMP_RET_KILL`` will always take precedence.)
In precedence order, they are: In precedence order, they are:
SECCOMP_RET_KILL: ``SECCOMP_RET_KILL``:
Results in the task exiting immediately without executing the Results in the task exiting immediately without executing the
system call. The exit status of the task (status & 0x7f) will system call. The exit status of the task (``status & 0x7f``) will
be SIGSYS, not SIGKILL. be ``SIGSYS``, not ``SIGKILL``.
SECCOMP_RET_TRAP: ``SECCOMP_RET_TRAP``:
Results in the kernel sending a SIGSYS signal to the triggering Results in the kernel sending a ``SIGSYS`` signal to the triggering
task without executing the system call. siginfo->si_call_addr task without executing the system call. ``siginfo->si_call_addr``
will show the address of the system call instruction, and will show the address of the system call instruction, and
siginfo->si_syscall and siginfo->si_arch will indicate which ``siginfo->si_syscall`` and ``siginfo->si_arch`` will indicate which
syscall was attempted. The program counter will be as though syscall was attempted. The program counter will be as though
the syscall happened (i.e. it will not point to the syscall the syscall happened (i.e. it will not point to the syscall
instruction). The return value register will contain an arch- instruction). The return value register will contain an arch-
dependent value -- if resuming execution, set it to something dependent value -- if resuming execution, set it to something
sensible. (The architecture dependency is because replacing sensible. (The architecture dependency is because replacing
it with -ENOSYS could overwrite some useful information.) it with ``-ENOSYS`` could overwrite some useful information.)
The SECCOMP_RET_DATA portion of the return value will be passed The ``SECCOMP_RET_DATA`` portion of the return value will be passed
as si_errno. as ``si_errno``.
SIGSYS triggered by seccomp will have a si_code of SYS_SECCOMP. ``SIGSYS`` triggered by seccomp will have a si_code of ``SYS_SECCOMP``.
SECCOMP_RET_ERRNO: ``SECCOMP_RET_ERRNO``:
Results in the lower 16-bits of the return value being passed Results in the lower 16-bits of the return value being passed
to userland as the errno without executing the system call. to userland as the errno without executing the system call.
SECCOMP_RET_TRACE: ``SECCOMP_RET_TRACE``:
When returned, this value will cause the kernel to attempt to When returned, this value will cause the kernel to attempt to
notify a ptrace()-based tracer prior to executing the system notify a ``ptrace()``-based tracer prior to executing the system
call. If there is no tracer present, -ENOSYS is returned to call. If there is no tracer present, ``-ENOSYS`` is returned to
userland and the system call is not executed. userland and the system call is not executed.
A tracer will be notified if it requests PTRACE_O_TRACESECCOMP A tracer will be notified if it requests ``PTRACE_O_TRACESECCOM``P
using ptrace(PTRACE_SETOPTIONS). The tracer will be notified using ``ptrace(PTRACE_SETOPTIONS)``. The tracer will be notified
of a PTRACE_EVENT_SECCOMP and the SECCOMP_RET_DATA portion of of a ``PTRACE_EVENT_SECCOMP`` and the ``SECCOMP_RET_DATA`` portion of
the BPF program return value will be available to the tracer the BPF program return value will be available to the tracer
via PTRACE_GETEVENTMSG. via ``PTRACE_GETEVENTMSG``.
The tracer can skip the system call by changing the syscall number The tracer can skip the system call by changing the syscall number
to -1. Alternatively, the tracer can change the system call to -1. Alternatively, the tracer can change the system call
...@@ -138,19 +141,19 @@ SECCOMP_RET_TRACE: ...@@ -138,19 +141,19 @@ SECCOMP_RET_TRACE:
allow use of ptrace, even of other sandboxed processes, without allow use of ptrace, even of other sandboxed processes, without
extreme care; ptracers can use this mechanism to escape.) extreme care; ptracers can use this mechanism to escape.)
SECCOMP_RET_ALLOW: ``SECCOMP_RET_ALLOW``:
Results in the system call being executed. Results in the system call being executed.
If multiple filters exist, the return value for the evaluation of a If multiple filters exist, the return value for the evaluation of a
given system call will always use the highest precedent value. given system call will always use the highest precedent value.
Precedence is only determined using the SECCOMP_RET_ACTION mask. When Precedence is only determined using the ``SECCOMP_RET_ACTION`` mask. When
multiple filters return values of the same precedence, only the multiple filters return values of the same precedence, only the
SECCOMP_RET_DATA from the most recently installed filter will be ``SECCOMP_RET_DATA`` from the most recently installed filter will be
returned. returned.
Pitfalls Pitfalls
-------- ========
The biggest pitfall to avoid during use is filtering on system call The biggest pitfall to avoid during use is filtering on system call
number without checking the architecture value. Why? On any number without checking the architecture value. Why? On any
...@@ -160,39 +163,40 @@ the numbers in the different calling conventions overlap, then checks in ...@@ -160,39 +163,40 @@ the numbers in the different calling conventions overlap, then checks in
the filters may be abused. Always check the arch value! the filters may be abused. Always check the arch value!
Example Example
------- =======
The samples/seccomp/ directory contains both an x86-specific example The ``samples/seccomp/`` directory contains both an x86-specific example
and a more generic example of a higher level macro interface for BPF and a more generic example of a higher level macro interface for BPF
program generation. program generation.
Adding architecture support Adding architecture support
----------------------- ===========================
See arch/Kconfig for the authoritative requirements. In general, if an See ``arch/Kconfig`` for the authoritative requirements. In general, if an
architecture supports both ptrace_event and seccomp, it will be able to architecture supports both ptrace_event and seccomp, it will be able to
support seccomp filter with minor fixup: SIGSYS support and seccomp return support seccomp filter with minor fixup: ``SIGSYS`` support and seccomp return
value checking. Then it must just add CONFIG_HAVE_ARCH_SECCOMP_FILTER value checking. Then it must just add ``CONFIG_HAVE_ARCH_SECCOMP_FILTER``
to its arch-specific Kconfig. to its arch-specific Kconfig.
Caveats Caveats
------- =======
The vDSO can cause some system calls to run entirely in userspace, The vDSO can cause some system calls to run entirely in userspace,
leading to surprises when you run programs on different machines that leading to surprises when you run programs on different machines that
fall back to real syscalls. To minimize these surprises on x86, make fall back to real syscalls. To minimize these surprises on x86, make
sure you test with sure you test with
/sys/devices/system/clocksource/clocksource0/current_clocksource set to ``/sys/devices/system/clocksource/clocksource0/current_clocksource`` set to
something like acpi_pm. something like ``acpi_pm``.
On x86-64, vsyscall emulation is enabled by default. (vsyscalls are On x86-64, vsyscall emulation is enabled by default. (vsyscalls are
legacy variants on vDSO calls.) Currently, emulated vsyscalls will honor seccomp, with a few oddities: legacy variants on vDSO calls.) Currently, emulated vsyscalls will
honor seccomp, with a few oddities:
- A return value of SECCOMP_RET_TRAP will set a si_call_addr pointing to - A return value of ``SECCOMP_RET_TRAP`` will set a ``si_call_addr`` pointing to
the vsyscall entry for the given call and not the address after the the vsyscall entry for the given call and not the address after the
'syscall' instruction. Any code which wants to restart the call 'syscall' instruction. Any code which wants to restart the call
should be aware that (a) a ret instruction has been emulated and (b) should be aware that (a) a ret instruction has been emulated and (b)
...@@ -200,7 +204,7 @@ legacy variants on vDSO calls.) Currently, emulated vsyscalls will honor seccom ...@@ -200,7 +204,7 @@ legacy variants on vDSO calls.) Currently, emulated vsyscalls will honor seccom
emulation security checks, making resuming the syscall mostly emulation security checks, making resuming the syscall mostly
pointless. pointless.
- A return value of SECCOMP_RET_TRACE will signal the tracer as usual, - A return value of ``SECCOMP_RET_TRACE`` will signal the tracer as usual,
but the syscall may not be changed to another system call using the but the syscall may not be changed to another system call using the
orig_rax register. It may only be changed to -1 order to skip the orig_rax register. It may only be changed to -1 order to skip the
currently emulated call. Any other change MAY terminate the process. currently emulated call. Any other change MAY terminate the process.
...@@ -209,14 +213,14 @@ legacy variants on vDSO calls.) Currently, emulated vsyscalls will honor seccom ...@@ -209,14 +213,14 @@ legacy variants on vDSO calls.) Currently, emulated vsyscalls will honor seccom
rip or rsp. (Do not rely on other changes terminating the process. rip or rsp. (Do not rely on other changes terminating the process.
They might work. For example, on some kernels, choosing a syscall They might work. For example, on some kernels, choosing a syscall
that only exists in future kernels will be correctly emulated (by that only exists in future kernels will be correctly emulated (by
returning -ENOSYS). returning ``-ENOSYS``).
To detect this quirky behavior, check for addr & ~0x0C00 == To detect this quirky behavior, check for ``addr & ~0x0C00 ==
0xFFFFFFFFFF600000. (For SECCOMP_RET_TRACE, use rip. For 0xFFFFFFFFFF600000``. (For ``SECCOMP_RET_TRACE``, use rip. For
SECCOMP_RET_TRAP, use siginfo->si_call_addr.) Do not check any other ``SECCOMP_RET_TRAP``, use ``siginfo->si_call_addr``.) Do not check any other
condition: future kernels may improve vsyscall emulation and current condition: future kernels may improve vsyscall emulation and current
kernels in vsyscall=native mode will behave differently, but the kernels in vsyscall=native mode will behave differently, but the
instructions at 0xF...F600{0,4,8,C}00 will not be system calls in these instructions at ``0xF...F600{0,4,8,C}00`` will not be system calls in these
cases. cases.
Note that modern systems are unlikely to use vsyscalls at all -- they Note that modern systems are unlikely to use vsyscalls at all -- they
......
...@@ -11492,6 +11492,7 @@ F: kernel/seccomp.c ...@@ -11492,6 +11492,7 @@ F: kernel/seccomp.c
F: include/uapi/linux/seccomp.h F: include/uapi/linux/seccomp.h
F: include/linux/seccomp.h F: include/linux/seccomp.h
F: tools/testing/selftests/seccomp/* F: tools/testing/selftests/seccomp/*
F: Documentation/userspace-api/seccomp_filter.rst
K: \bsecure_computing K: \bsecure_computing
K: \bTIF_SECCOMP\b K: \bTIF_SECCOMP\b
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment