lkml.org 
[lkml]   [2012]   [Apr]   [17]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH v18 15/15] Documentation: prctl/seccomp_filter
From
On Tue, Apr 17, 2012 at 9:28 PM, Paul Gortmaker
<paul.gortmaker@windriver.com> wrote:
> On Thu, Apr 12, 2012 at 5:48 PM, Will Drewry <wad@chromium.org> wrote:
>> Documents how system call filtering using Berkeley Packet
>> Filter programs works and how it may be used.
>> Includes an example for x86 and a semi-generic
>> example using a macro-based code generator.
>
> Hi guys, that is quite the To/CC list.  I didn't have the heart to add
> yet another address (linux-next) to it...
>
> Can you have a look at this link?   It appears to be breaking
> the i386 allmodconfig builds in linux-next.
>
> http://kisskb.ellerman.id.au/kisskb/buildresult/6123842/
>
>    "samples/seccomp/bpf-direct.c:47:2: error: #error Unsupported platform"

Ouch. It's checking the __i386__ and __x86_64__ macros. I'll reproduce
it and see what needs to change. I suspect it'll be something dumb,
but the worst case is that I drop that bit of fanciness.

Thanks and sorry!
will

>>
>> Acked-by: Eric Paris <eparis@redhat.com>
>> Signed-off-by: Will Drewry <wad@chromium.org>
>>
>> v18: - added acked by
>>     - update no new privs numbers
>> v17: - remove @compat note and add Pitfalls section for arch checking
>>       (keescook@chromium.org)
>> v16: -
>> v15: -
>> v14: - rebase/nochanges
>> v13: - rebase on to 88ebdda6159ffc15699f204c33feb3e431bf9bdc
>> v12: - comment on the ptrace_event use
>>     - update arch support comment
>>     - note the behavior of SECCOMP_RET_DATA when there are multiple filters
>>       (keescook@chromium.org)
>>     - lots of samples/ clean up incl 64-bit bpf-direct support
>>       (markus@chromium.org)
>>     - rebase to linux-next
>> v11: - overhaul return value language, updates (keescook@chromium.org)
>>     - comment on do_exit(SIGSYS)
>> v10: - update for SIGSYS
>>     - update for new seccomp_data layout
>>     - update for ptrace option use
>> v9: - updated bpf-direct.c for SIGILL
>> v8: - add PR_SET_NO_NEW_PRIVS to the samples.
>> v7: - updated for all the new stuff in v7: TRAP, TRACE
>>    - only talk about PR_SET_SECCOMP now
>>    - fixed bad JLE32 check (coreyb@linux.vnet.ibm.com)
>>    - adds dropper.c: a simple system call disabler
>> v6: - tweak the language to note the requirement of
>>      PR_SET_NO_NEW_PRIVS being called prior to use. (luto@mit.edu)
>> v5: - update sample to use system call arguments
>>    - adds a "fancy" example using a macro-based generator
>>    - cleaned up bpf in the sample
>>    - update docs to mention arguments
>>    - fix prctl value (eparis@redhat.com)
>>    - language cleanup (rdunlap@xenotime.net)
>> v4: - update for no_new_privs use
>>    - minor tweaks
>> v3: - call out BPF <-> Berkeley Packet Filter (rdunlap@xenotime.net)
>>    - document use of tentative always-unprivileged
>>    - guard sample compilation for i386 and x86_64
>> v2: - move code to samples (corbet@lwn.net)
>> ---
>>  Documentation/prctl/seccomp_filter.txt |  163 ++++++++++++++++++++++
>>  samples/Makefile                       |    2 +-
>>  samples/seccomp/Makefile               |   38 +++++
>>  samples/seccomp/bpf-direct.c           |  176 +++++++++++++++++++++++
>>  samples/seccomp/bpf-fancy.c            |  102 ++++++++++++++
>>  samples/seccomp/bpf-helper.c           |   89 ++++++++++++
>>  samples/seccomp/bpf-helper.h           |  238 ++++++++++++++++++++++++++++++++
>>  samples/seccomp/dropper.c              |   68 +++++++++
>>  8 files changed, 875 insertions(+), 1 deletions(-)
>>  create mode 100644 Documentation/prctl/seccomp_filter.txt
>>  create mode 100644 samples/seccomp/Makefile
>>  create mode 100644 samples/seccomp/bpf-direct.c
>>  create mode 100644 samples/seccomp/bpf-fancy.c
>>  create mode 100644 samples/seccomp/bpf-helper.c
>>  create mode 100644 samples/seccomp/bpf-helper.h
>>  create mode 100644 samples/seccomp/dropper.c
>>
>> diff --git a/Documentation/prctl/seccomp_filter.txt b/Documentation/prctl/seccomp_filter.txt
>> new file mode 100644
>> index 0000000..597c3c5
>> --- /dev/null
>> +++ b/Documentation/prctl/seccomp_filter.txt
>> @@ -0,0 +1,163 @@
>> +               SECure COMPuting with filters
>> +               =============================
>> +
>> +Introduction
>> +------------
>> +
>> +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.
>> +As system calls change and mature, bugs are found and eradicated.  A
>> +certain subset of userland applications benefit by having a reduced set
>> +of available system calls.  The resulting set reduces the total kernel
>> +surface exposed to the application.  System call filtering is meant for
>> +use with those applications.
>> +
>> +Seccomp filtering provides a means for a process to specify a filter for
>> +incoming system calls.  The filter is expressed as a Berkeley Packet
>> +Filter (BPF) program, as with socket filters, except that the data
>> +operated on is related to the system call being made: system call
>> +number and the system call arguments.  This allows for expressive
>> +filtering of system calls using a filter program language with a long
>> +history of being exposed to userland and a straightforward data set.
>> +
>> +Additionally, BPF makes it impossible for users of seccomp to fall prey
>> +to time-of-check-time-of-use (TOCTOU) attacks that are common in system
>> +call interposition frameworks.  BPF programs may not dereference
>> +pointers which constrains all filters to solely evaluating the system
>> +call arguments directly.
>> +
>> +What it isn't
>> +-------------
>> +
>> +System call filtering isn't a sandbox.  It provides a clearly defined
>> +mechanism for minimizing the exposed kernel surface.  It is meant to be
>> +a tool for sandbox developers to use.  Beyond that, policy for logical
>> +behavior and information flow should be managed with a combination of
>> +other system hardening techniques and, potentially, an LSM of your
>> +choosing.  Expressive, dynamic filters provide further options down this
>> +path (avoiding pathological sizes or selecting which of the multiplexed
>> +system calls in socketcall() is allowed, for instance) which could be
>> +construed, incorrectly, as a more complete sandboxing solution.
>> +
>> +Usage
>> +-----
>> +
>> +An additional seccomp mode is added and is enabled using the same
>> +prctl(2) call as the strict seccomp.  If the architecture has
>> +CONFIG_HAVE_ARCH_SECCOMP_FILTER, then filters may be added as below:
>> +
>> +PR_SET_SECCOMP:
>> +       Now takes an additional argument which specifies a new filter
>> +       using a BPF program.
>> +       The BPF program will be executed over struct seccomp_data
>> +       reflecting the system call number, arguments, and other
>> +       metadata.  The BPF program must then return one of the
>> +       acceptable values to inform the kernel which action should be
>> +       taken.
>> +
>> +       Usage:
>> +               prctl(PR_SET_SECCOMP, SECCOMP_MODE_FILTER, prog);
>> +
>> +       The 'prog' argument is a pointer to a struct sock_fprog which
>> +       will contain the filter program.  If the program is invalid, the
>> +       call will return -1 and set errno to EINVAL.
>> +
>> +       If fork/clone and execve are allowed by @prog, any child
>> +       processes will be constrained to the same filters and system
>> +       call ABI as the parent.
>> +
>> +       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
>> +       true, -EACCES will be returned.  This requirement ensures that filter
>> +       programs cannot be applied to child processes with greater privileges
>> +       than the task that installed them.
>> +
>> +       Additionally, if prctl(2) is allowed by the attached filter,
>> +       additional filters may be layered on which will increase evaluation
>> +       time, but allow for further decreasing the attack surface during
>> +       execution of a process.
>> +
>> +The above call returns 0 on success and non-zero on error.
>> +
>> +Return values
>> +-------------
>> +A seccomp filter may return any of the following values. If multiple
>> +filters exist, the return value for the evaluation of a given system
>> +call will always use the highest precedent value. (For example,
>> +SECCOMP_RET_KILL will always take precedence.)
>> +
>> +In precedence order, they are:
>> +
>> +SECCOMP_RET_KILL:
>> +       Results in the task exiting immediately without executing the
>> +       system call.  The exit status of the task (status & 0x7f) will
>> +       be SIGSYS, not SIGKILL.
>> +
>> +SECCOMP_RET_TRAP:
>> +       Results in the kernel sending a SIGSYS signal to the triggering
>> +       task without executing the system call.  The kernel will
>> +       rollback the register state to just before the system call
>> +       entry such that a signal handler in the task will be able to
>> +       inspect the ucontext_t->uc_mcontext registers and emulate
>> +       system call success or failure upon return from the signal
>> +       handler.
>> +
>> +       The SECCOMP_RET_DATA portion of the return value will be passed
>> +       as si_errno.
>> +
>> +       SIGSYS triggered by seccomp will have a si_code of SYS_SECCOMP.
>> +
>> +SECCOMP_RET_ERRNO:
>> +       Results in the lower 16-bits of the return value being passed
>> +       to userland as the errno without executing the system call.
>> +
>> +SECCOMP_RET_TRACE:
>> +       When returned, this value will cause the kernel to attempt to
>> +       notify a ptrace()-based tracer prior to executing the system
>> +       call.  If there is no tracer present, -ENOSYS is returned to
>> +       userland and the system call is not executed.
>> +
>> +       A tracer will be notified if it requests PTRACE_O_TRACESECCOMP
>> +       using ptrace(PTRACE_SETOPTIONS).  The tracer will be notified
>> +       of a PTRACE_EVENT_SECCOMP and the SECCOMP_RET_DATA portion of
>> +       the BPF program return value will be available to the tracer
>> +       via PTRACE_GETEVENTMSG.
>> +
>> +SECCOMP_RET_ALLOW:
>> +       Results in the system call being executed.
>> +
>> +If multiple filters exist, the return value for the evaluation of a
>> +given system call will always use the highest precedent value.
>> +
>> +Precedence is only determined using the SECCOMP_RET_ACTION mask.  When
>> +multiple filters return values of the same precedence, only the
>> +SECCOMP_RET_DATA from the most recently installed filter will be
>> +returned.
>> +
>> +Pitfalls
>> +--------
>> +
>> +The biggest pitfall to avoid during use is filtering on system call
>> +number without checking the architecture value.  Why?  On any
>> +architecture that supports multiple system call invocation conventions,
>> +the system call numbers may vary based on the specific invocation.  If
>> +the numbers in the different calling conventions overlap, then checks in
>> +the filters may be abused.  Always check the arch value!
>> +
>> +Example
>> +-------
>> +
>> +The samples/seccomp/ directory contains both an x86-specific example
>> +and a more generic example of a higher level macro interface for BPF
>> +program generation.
>> +
>> +
>> +
>> +Adding architecture support
>> +-----------------------
>> +
>> +See arch/Kconfig for the authoritative requirements.  In general, if an
>> +architecture supports both ptrace_event and seccomp, it will be able to
>> +support seccomp filter with minor fixup: SIGSYS support and seccomp return
>> +value checking.  Then it must just add CONFIG_HAVE_ARCH_SECCOMP_FILTER
>> +to its arch-specific Kconfig.
>> diff --git a/samples/Makefile b/samples/Makefile
>> index 2f75851..5ef08bb 100644
>> --- a/samples/Makefile
>> +++ b/samples/Makefile
>> @@ -1,4 +1,4 @@
>>  # Makefile for Linux samples code
>>
>>  obj-$(CONFIG_SAMPLES)  += kobject/ kprobes/ tracepoints/ trace_events/ \
>> -                          hw_breakpoint/ kfifo/ kdb/ hidraw/ rpmsg/
>> +                          hw_breakpoint/ kfifo/ kdb/ hidraw/ rpmsg/ seccomp/
>> diff --git a/samples/seccomp/Makefile b/samples/seccomp/Makefile
>> new file mode 100644
>> index 0000000..e8fe0f5
>> --- /dev/null
>> +++ b/samples/seccomp/Makefile
>> @@ -0,0 +1,38 @@
>> +# kbuild trick to avoid linker error. Can be omitted if a module is built.
>> +obj- := dummy.o
>> +
>> +hostprogs-$(CONFIG_SECCOMP) := bpf-fancy dropper
>> +bpf-fancy-objs := bpf-fancy.o bpf-helper.o
>> +
>> +HOSTCFLAGS_bpf-fancy.o += -I$(objtree)/usr/include
>> +HOSTCFLAGS_bpf-fancy.o += -idirafter $(objtree)/include
>> +HOSTCFLAGS_bpf-helper.o += -I$(objtree)/usr/include
>> +HOSTCFLAGS_bpf-helper.o += -idirafter $(objtree)/include
>> +
>> +HOSTCFLAGS_dropper.o += -I$(objtree)/usr/include
>> +HOSTCFLAGS_dropper.o += -idirafter $(objtree)/include
>> +dropper-objs := dropper.o
>> +
>> +# bpf-direct.c is x86-only.
>> +ifeq ($(SRCARCH),x86)
>> +# List of programs to build
>> +hostprogs-$(CONFIG_SECCOMP) += bpf-direct
>> +bpf-direct-objs := bpf-direct.o
>> +endif
>> +
>> +HOSTCFLAGS_bpf-direct.o += -I$(objtree)/usr/include
>> +HOSTCFLAGS_bpf-direct.o += -idirafter $(objtree)/include
>> +
>> +# Try to match the kernel target.
>> +ifeq ($(CONFIG_64BIT),)
>> +HOSTCFLAGS_bpf-direct.o += -m32
>> +HOSTCFLAGS_dropper.o += -m32
>> +HOSTCFLAGS_bpf-helper.o += -m32
>> +HOSTCFLAGS_bpf-fancy.o += -m32
>> +HOSTLOADLIBES_bpf-direct += -m32
>> +HOSTLOADLIBES_bpf-fancy += -m32
>> +HOSTLOADLIBES_dropper += -m32
>> +endif
>> +
>> +# Tell kbuild to always build the programs
>> +always := $(hostprogs-y)
>> diff --git a/samples/seccomp/bpf-direct.c b/samples/seccomp/bpf-direct.c
>> new file mode 100644
>> index 0000000..26f523e
>> --- /dev/null
>> +++ b/samples/seccomp/bpf-direct.c
>> @@ -0,0 +1,176 @@
>> +/*
>> + * Seccomp filter example for x86 (32-bit and 64-bit) with BPF macros
>> + *
>> + * Copyright (c) 2012 The Chromium OS Authors <chromium-os-dev@chromium.org>
>> + * Author: Will Drewry <wad@chromium.org>
>> + *
>> + * The code may be used by anyone for any purpose,
>> + * and can serve as a starting point for developing
>> + * applications using prctl(PR_SET_SECCOMP, 2, ...).
>> + */
>> +#define __USE_GNU 1
>> +#define _GNU_SOURCE 1
>> +
>> +#include <linux/types.h>
>> +#include <linux/filter.h>
>> +#include <linux/seccomp.h>
>> +#include <linux/unistd.h>
>> +#include <signal.h>
>> +#include <stdio.h>
>> +#include <stddef.h>
>> +#include <string.h>
>> +#include <sys/prctl.h>
>> +#include <unistd.h>
>> +
>> +#define syscall_arg(_n) (offsetof(struct seccomp_data, args[_n]))
>> +#define syscall_nr (offsetof(struct seccomp_data, nr))
>> +
>> +#if defined(__i386__)
>> +#define REG_RESULT     REG_EAX
>> +#define REG_SYSCALL    REG_EAX
>> +#define REG_ARG0       REG_EBX
>> +#define REG_ARG1       REG_ECX
>> +#define REG_ARG2       REG_EDX
>> +#define REG_ARG3       REG_ESI
>> +#define REG_ARG4       REG_EDI
>> +#define REG_ARG5       REG_EBP
>> +#elif defined(__x86_64__)
>> +#define REG_RESULT     REG_RAX
>> +#define REG_SYSCALL    REG_RAX
>> +#define REG_ARG0       REG_RDI
>> +#define REG_ARG1       REG_RSI
>> +#define REG_ARG2       REG_RDX
>> +#define REG_ARG3       REG_R10
>> +#define REG_ARG4       REG_R8
>> +#define REG_ARG5       REG_R9
>> +#else
>> +#error Unsupported platform
>> +#endif
>> +
>> +#ifndef PR_SET_NO_NEW_PRIVS
>> +#define PR_SET_NO_NEW_PRIVS 38
>> +#endif
>> +
>> +#ifndef SYS_SECCOMP
>> +#define SYS_SECCOMP 1
>> +#endif
>> +
>> +static void emulator(int nr, siginfo_t *info, void *void_context)
>> +{
>> +       ucontext_t *ctx = (ucontext_t *)(void_context);
>> +       int syscall;
>> +       char *buf;
>> +       ssize_t bytes;
>> +       size_t len;
>> +       if (info->si_code != SYS_SECCOMP)
>> +               return;
>> +       if (!ctx)
>> +               return;
>> +       syscall = ctx->uc_mcontext.gregs[REG_SYSCALL];
>> +       buf = (char *) ctx->uc_mcontext.gregs[REG_ARG1];
>> +       len = (size_t) ctx->uc_mcontext.gregs[REG_ARG2];
>> +
>> +       if (syscall != __NR_write)
>> +               return;
>> +       if (ctx->uc_mcontext.gregs[REG_ARG0] != STDERR_FILENO)
>> +               return;
>> +       /* Redirect stderr messages to stdout. Doesn't handle EINTR, etc */
>> +       ctx->uc_mcontext.gregs[REG_RESULT] = -1;
>> +       if (write(STDOUT_FILENO, "[ERR] ", 6) > 0) {
>> +               bytes = write(STDOUT_FILENO, buf, len);
>> +               ctx->uc_mcontext.gregs[REG_RESULT] = bytes;
>> +       }
>> +       return;
>> +}
>> +
>> +static int install_emulator(void)
>> +{
>> +       struct sigaction act;
>> +       sigset_t mask;
>> +       memset(&act, 0, sizeof(act));
>> +       sigemptyset(&mask);
>> +       sigaddset(&mask, SIGSYS);
>> +
>> +       act.sa_sigaction = &emulator;
>> +       act.sa_flags = SA_SIGINFO;
>> +       if (sigaction(SIGSYS, &act, NULL) < 0) {
>> +               perror("sigaction");
>> +               return -1;
>> +       }
>> +       if (sigprocmask(SIG_UNBLOCK, &mask, NULL)) {
>> +               perror("sigprocmask");
>> +               return -1;
>> +       }
>> +       return 0;
>> +}
>> +
>> +static int install_filter(void)
>> +{
>> +       struct sock_filter filter[] = {
>> +               /* Grab the system call number */
>> +               BPF_STMT(BPF_LD+BPF_W+BPF_ABS, syscall_nr),
>> +               /* Jump table for the allowed syscalls */
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, __NR_rt_sigreturn, 0, 1),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_ALLOW),
>> +#ifdef __NR_sigreturn
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, __NR_sigreturn, 0, 1),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_ALLOW),
>> +#endif
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, __NR_exit_group, 0, 1),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_ALLOW),
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, __NR_exit, 0, 1),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_ALLOW),
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, __NR_read, 1, 0),
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, __NR_write, 3, 2),
>> +
>> +               /* Check that read is only using stdin. */
>> +               BPF_STMT(BPF_LD+BPF_W+BPF_ABS, syscall_arg(0)),
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, STDIN_FILENO, 4, 0),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_KILL),
>> +
>> +               /* Check that write is only using stdout */
>> +               BPF_STMT(BPF_LD+BPF_W+BPF_ABS, syscall_arg(0)),
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, STDOUT_FILENO, 1, 0),
>> +               /* Trap attempts to write to stderr */
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, STDERR_FILENO, 1, 2),
>> +
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_ALLOW),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_TRAP),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_KILL),
>> +       };
>> +       struct sock_fprog prog = {
>> +               .len = (unsigned short)(sizeof(filter)/sizeof(filter[0])),
>> +               .filter = filter,
>> +       };
>> +
>> +       if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) {
>> +               perror("prctl(NO_NEW_PRIVS)");
>> +               return 1;
>> +       }
>> +
>> +
>> +       if (prctl(PR_SET_SECCOMP, SECCOMP_MODE_FILTER, &prog)) {
>> +               perror("prctl");
>> +               return 1;
>> +       }
>> +       return 0;
>> +}
>> +
>> +#define payload(_c) (_c), sizeof((_c))
>> +int main(int argc, char **argv)
>> +{
>> +       char buf[4096];
>> +       ssize_t bytes = 0;
>> +       if (install_emulator())
>> +               return 1;
>> +       if (install_filter())
>> +               return 1;
>> +       syscall(__NR_write, STDOUT_FILENO,
>> +               payload("OHAI! WHAT IS YOUR NAME? "));
>> +       bytes = syscall(__NR_read, STDIN_FILENO, buf, sizeof(buf));
>> +       syscall(__NR_write, STDOUT_FILENO, payload("HELLO, "));
>> +       syscall(__NR_write, STDOUT_FILENO, buf, bytes);
>> +       syscall(__NR_write, STDERR_FILENO,
>> +               payload("Error message going to STDERR\n"));
>> +       return 0;
>> +}
>> diff --git a/samples/seccomp/bpf-fancy.c b/samples/seccomp/bpf-fancy.c
>> new file mode 100644
>> index 0000000..8eb483a
>> --- /dev/null
>> +++ b/samples/seccomp/bpf-fancy.c
>> @@ -0,0 +1,102 @@
>> +/*
>> + * Seccomp BPF example using a macro-based generator.
>> + *
>> + * Copyright (c) 2012 The Chromium OS Authors <chromium-os-dev@chromium.org>
>> + * Author: Will Drewry <wad@chromium.org>
>> + *
>> + * The code may be used by anyone for any purpose,
>> + * and can serve as a starting point for developing
>> + * applications using prctl(PR_ATTACH_SECCOMP_FILTER).
>> + */
>> +
>> +#include <linux/filter.h>
>> +#include <linux/seccomp.h>
>> +#include <linux/unistd.h>
>> +#include <stdio.h>
>> +#include <string.h>
>> +#include <sys/prctl.h>
>> +#include <unistd.h>
>> +
>> +#include "bpf-helper.h"
>> +
>> +#ifndef PR_SET_NO_NEW_PRIVS
>> +#define PR_SET_NO_NEW_PRIVS 38
>> +#endif
>> +
>> +int main(int argc, char **argv)
>> +{
>> +       struct bpf_labels l;
>> +       static const char msg1[] = "Please type something: ";
>> +       static const char msg2[] = "You typed: ";
>> +       char buf[256];
>> +       struct sock_filter filter[] = {
>> +               /* TODO: LOAD_SYSCALL_NR(arch) and enforce an arch */
>> +               LOAD_SYSCALL_NR,
>> +               SYSCALL(__NR_exit, ALLOW),
>> +               SYSCALL(__NR_exit_group, ALLOW),
>> +               SYSCALL(__NR_write, JUMP(&l, write_fd)),
>> +               SYSCALL(__NR_read, JUMP(&l, read)),
>> +               DENY,  /* Don't passthrough into a label */
>> +
>> +               LABEL(&l, read),
>> +               ARG(0),
>> +               JNE(STDIN_FILENO, DENY),
>> +               ARG(1),
>> +               JNE((unsigned long)buf, DENY),
>> +               ARG(2),
>> +               JGE(sizeof(buf), DENY),
>> +               ALLOW,
>> +
>> +               LABEL(&l, write_fd),
>> +               ARG(0),
>> +               JEQ(STDOUT_FILENO, JUMP(&l, write_buf)),
>> +               JEQ(STDERR_FILENO, JUMP(&l, write_buf)),
>> +               DENY,
>> +
>> +               LABEL(&l, write_buf),
>> +               ARG(1),
>> +               JEQ((unsigned long)msg1, JUMP(&l, msg1_len)),
>> +               JEQ((unsigned long)msg2, JUMP(&l, msg2_len)),
>> +               JEQ((unsigned long)buf, JUMP(&l, buf_len)),
>> +               DENY,
>> +
>> +               LABEL(&l, msg1_len),
>> +               ARG(2),
>> +               JLT(sizeof(msg1), ALLOW),
>> +               DENY,
>> +
>> +               LABEL(&l, msg2_len),
>> +               ARG(2),
>> +               JLT(sizeof(msg2), ALLOW),
>> +               DENY,
>> +
>> +               LABEL(&l, buf_len),
>> +               ARG(2),
>> +               JLT(sizeof(buf), ALLOW),
>> +               DENY,
>> +       };
>> +       struct sock_fprog prog = {
>> +               .filter = filter,
>> +               .len = (unsigned short)(sizeof(filter)/sizeof(filter[0])),
>> +       };
>> +       ssize_t bytes;
>> +       bpf_resolve_jumps(&l, filter, sizeof(filter)/sizeof(*filter));
>> +
>> +       if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) {
>> +               perror("prctl(NO_NEW_PRIVS)");
>> +               return 1;
>> +       }
>> +
>> +       if (prctl(PR_SET_SECCOMP, SECCOMP_MODE_FILTER, &prog)) {
>> +               perror("prctl(SECCOMP)");
>> +               return 1;
>> +       }
>> +       syscall(__NR_write, STDOUT_FILENO, msg1, strlen(msg1));
>> +       bytes = syscall(__NR_read, STDIN_FILENO, buf, sizeof(buf)-1);
>> +       bytes = (bytes > 0 ? bytes : 0);
>> +       syscall(__NR_write, STDERR_FILENO, msg2, strlen(msg2));
>> +       syscall(__NR_write, STDERR_FILENO, buf, bytes);
>> +       /* Now get killed */
>> +       syscall(__NR_write, STDERR_FILENO, msg2, strlen(msg2)+2);
>> +       return 0;
>> +}
>> diff --git a/samples/seccomp/bpf-helper.c b/samples/seccomp/bpf-helper.c
>> new file mode 100644
>> index 0000000..579cfe3
>> --- /dev/null
>> +++ b/samples/seccomp/bpf-helper.c
>> @@ -0,0 +1,89 @@
>> +/*
>> + * Seccomp BPF helper functions
>> + *
>> + * Copyright (c) 2012 The Chromium OS Authors <chromium-os-dev@chromium.org>
>> + * Author: Will Drewry <wad@chromium.org>
>> + *
>> + * The code may be used by anyone for any purpose,
>> + * and can serve as a starting point for developing
>> + * applications using prctl(PR_ATTACH_SECCOMP_FILTER).
>> + */
>> +
>> +#include <stdio.h>
>> +#include <string.h>
>> +
>> +#include "bpf-helper.h"
>> +
>> +int bpf_resolve_jumps(struct bpf_labels *labels,
>> +                     struct sock_filter *filter, size_t count)
>> +{
>> +       struct sock_filter *begin = filter;
>> +       __u8 insn = count - 1;
>> +
>> +       if (count < 1)
>> +               return -1;
>> +       /*
>> +       * Walk it once, backwards, to build the label table and do fixups.
>> +       * Since backward jumps are disallowed by BPF, this is easy.
>> +       */
>> +       filter += insn;
>> +       for (; filter >= begin; --insn, --filter) {
>> +               if (filter->code != (BPF_JMP+BPF_JA))
>> +                       continue;
>> +               switch ((filter->jt<<8)|filter->jf) {
>> +               case (JUMP_JT<<8)|JUMP_JF:
>> +                       if (labels->labels[filter->k].location == 0xffffffff) {
>> +                               fprintf(stderr, "Unresolved label: '%s'\n",
>> +                                       labels->labels[filter->k].label);
>> +                               return 1;
>> +                       }
>> +                       filter->k = labels->labels[filter->k].location -
>> +                                   (insn + 1);
>> +                       filter->jt = 0;
>> +                       filter->jf = 0;
>> +                       continue;
>> +               case (LABEL_JT<<8)|LABEL_JF:
>> +                       if (labels->labels[filter->k].location != 0xffffffff) {
>> +                               fprintf(stderr, "Duplicate label use: '%s'\n",
>> +                                       labels->labels[filter->k].label);
>> +                               return 1;
>> +                       }
>> +                       labels->labels[filter->k].location = insn;
>> +                       filter->k = 0; /* fall through */
>> +                       filter->jt = 0;
>> +                       filter->jf = 0;
>> +                       continue;
>> +               }
>> +       }
>> +       return 0;
>> +}
>> +
>> +/* Simple lookup table for labels. */
>> +__u32 seccomp_bpf_label(struct bpf_labels *labels, const char *label)
>> +{
>> +       struct __bpf_label *begin = labels->labels, *end;
>> +       int id;
>> +       if (labels->count == 0) {
>> +               begin->label = label;
>> +               begin->location = 0xffffffff;
>> +               labels->count++;
>> +               return 0;
>> +       }
>> +       end = begin + labels->count;
>> +       for (id = 0; begin < end; ++begin, ++id) {
>> +               if (!strcmp(label, begin->label))
>> +                       return id;
>> +       }
>> +       begin->label = label;
>> +       begin->location = 0xffffffff;
>> +       labels->count++;
>> +       return id;
>> +}
>> +
>> +void seccomp_bpf_print(struct sock_filter *filter, size_t count)
>> +{
>> +       struct sock_filter *end = filter + count;
>> +       for ( ; filter < end; ++filter)
>> +               printf("{ code=%u,jt=%u,jf=%u,k=%u },\n",
>> +                       filter->code, filter->jt, filter->jf, filter->k);
>> +}
>> diff --git a/samples/seccomp/bpf-helper.h b/samples/seccomp/bpf-helper.h
>> new file mode 100644
>> index 0000000..643279d
>> --- /dev/null
>> +++ b/samples/seccomp/bpf-helper.h
>> @@ -0,0 +1,238 @@
>> +/*
>> + * Example wrapper around BPF macros.
>> + *
>> + * Copyright (c) 2012 The Chromium OS Authors <chromium-os-dev@chromium.org>
>> + * Author: Will Drewry <wad@chromium.org>
>> + *
>> + * The code may be used by anyone for any purpose,
>> + * and can serve as a starting point for developing
>> + * applications using prctl(PR_SET_SECCOMP, 2, ...).
>> + *
>> + * No guarantees are provided with respect to the correctness
>> + * or functionality of this code.
>> + */
>> +#ifndef __BPF_HELPER_H__
>> +#define __BPF_HELPER_H__
>> +
>> +#include <asm/bitsperlong.h>   /* for __BITS_PER_LONG */
>> +#include <endian.h>
>> +#include <linux/filter.h>
>> +#include <linux/seccomp.h>     /* for seccomp_data */
>> +#include <linux/types.h>
>> +#include <linux/unistd.h>
>> +#include <stddef.h>
>> +
>> +#define BPF_LABELS_MAX 256
>> +struct bpf_labels {
>> +       int count;
>> +       struct __bpf_label {
>> +               const char *label;
>> +               __u32 location;
>> +       } labels[BPF_LABELS_MAX];
>> +};
>> +
>> +int bpf_resolve_jumps(struct bpf_labels *labels,
>> +                     struct sock_filter *filter, size_t count);
>> +__u32 seccomp_bpf_label(struct bpf_labels *labels, const char *label);
>> +void seccomp_bpf_print(struct sock_filter *filter, size_t count);
>> +
>> +#define JUMP_JT 0xff
>> +#define JUMP_JF 0xff
>> +#define LABEL_JT 0xfe
>> +#define LABEL_JF 0xfe
>> +
>> +#define ALLOW \
>> +       BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_ALLOW)
>> +#define DENY \
>> +       BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_KILL)
>> +#define JUMP(labels, label) \
>> +       BPF_JUMP(BPF_JMP+BPF_JA, FIND_LABEL((labels), (label)), \
>> +                JUMP_JT, JUMP_JF)
>> +#define LABEL(labels, label) \
>> +       BPF_JUMP(BPF_JMP+BPF_JA, FIND_LABEL((labels), (label)), \
>> +                LABEL_JT, LABEL_JF)
>> +#define SYSCALL(nr, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (nr), 0, 1), \
>> +       jt
>> +
>> +/* Lame, but just an example */
>> +#define FIND_LABEL(labels, label) seccomp_bpf_label((labels), #label)
>> +
>> +#define EXPAND(...) __VA_ARGS__
>> +/* Map all width-sensitive operations */
>> +#if __BITS_PER_LONG == 32
>> +
>> +#define JEQ(x, jt) JEQ32(x, EXPAND(jt))
>> +#define JNE(x, jt) JNE32(x, EXPAND(jt))
>> +#define JGT(x, jt) JGT32(x, EXPAND(jt))
>> +#define JLT(x, jt) JLT32(x, EXPAND(jt))
>> +#define JGE(x, jt) JGE32(x, EXPAND(jt))
>> +#define JLE(x, jt) JLE32(x, EXPAND(jt))
>> +#define JA(x, jt) JA32(x, EXPAND(jt))
>> +#define ARG(i) ARG_32(i)
>> +#define LO_ARG(idx) offsetof(struct seccomp_data, args[(idx)])
>> +
>> +#elif __BITS_PER_LONG == 64
>> +
>> +/* Ensure that we load the logically correct offset. */
>> +#if __BYTE_ORDER == __LITTLE_ENDIAN
>> +#define ENDIAN(_lo, _hi) _lo, _hi
>> +#define LO_ARG(idx) offsetof(struct seccomp_data, args[(idx)])
>> +#define HI_ARG(idx) offsetof(struct seccomp_data, args[(idx)]) + sizeof(__u32)
>> +#elif __BYTE_ORDER == __BIG_ENDIAN
>> +#define ENDIAN(_lo, _hi) _hi, _lo
>> +#define LO_ARG(idx) offsetof(struct seccomp_data, args[(idx)]) + sizeof(__u32)
>> +#define HI_ARG(idx) offsetof(struct seccomp_data, args[(idx)])
>> +#else
>> +#error "Unknown endianness"
>> +#endif
>> +
>> +union arg64 {
>> +       struct {
>> +               __u32 ENDIAN(lo32, hi32);
>> +       };
>> +       __u64 u64;
>> +};
>> +
>> +#define JEQ(x, jt) \
>> +       JEQ64(((union arg64){.u64 = (x)}).lo32, \
>> +             ((union arg64){.u64 = (x)}).hi32, \
>> +             EXPAND(jt))
>> +#define JGT(x, jt) \
>> +       JGT64(((union arg64){.u64 = (x)}).lo32, \
>> +             ((union arg64){.u64 = (x)}).hi32, \
>> +             EXPAND(jt))
>> +#define JGE(x, jt) \
>> +       JGE64(((union arg64){.u64 = (x)}).lo32, \
>> +             ((union arg64){.u64 = (x)}).hi32, \
>> +             EXPAND(jt))
>> +#define JNE(x, jt) \
>> +       JNE64(((union arg64){.u64 = (x)}).lo32, \
>> +             ((union arg64){.u64 = (x)}).hi32, \
>> +             EXPAND(jt))
>> +#define JLT(x, jt) \
>> +       JLT64(((union arg64){.u64 = (x)}).lo32, \
>> +             ((union arg64){.u64 = (x)}).hi32, \
>> +             EXPAND(jt))
>> +#define JLE(x, jt) \
>> +       JLE64(((union arg64){.u64 = (x)}).lo32, \
>> +             ((union arg64){.u64 = (x)}).hi32, \
>> +             EXPAND(jt))
>> +
>> +#define JA(x, jt) \
>> +       JA64(((union arg64){.u64 = (x)}).lo32, \
>> +              ((union arg64){.u64 = (x)}).hi32, \
>> +              EXPAND(jt))
>> +#define ARG(i) ARG_64(i)
>> +
>> +#else
>> +#error __BITS_PER_LONG value unusable.
>> +#endif
>> +
>> +/* Loads the arg into A */
>> +#define ARG_32(idx) \
>> +       BPF_STMT(BPF_LD+BPF_W+BPF_ABS, LO_ARG(idx))
>> +
>> +/* Loads hi into A and lo in X */
>> +#define ARG_64(idx) \
>> +       BPF_STMT(BPF_LD+BPF_W+BPF_ABS, LO_ARG(idx)), \
>> +       BPF_STMT(BPF_ST, 0), /* lo -> M[0] */ \
>> +       BPF_STMT(BPF_LD+BPF_W+BPF_ABS, HI_ARG(idx)), \
>> +       BPF_STMT(BPF_ST, 1) /* hi -> M[1] */
>> +
>> +#define JEQ32(value, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (value), 0, 1), \
>> +       jt
>> +
>> +#define JNE32(value, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (value), 1, 0), \
>> +       jt
>> +
>> +/* Checks the lo, then swaps to check the hi. A=lo,X=hi */
>> +#define JEQ64(lo, hi, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (hi), 0, 5), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 0), /* swap in lo */ \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (lo), 0, 2), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1), /* passed: swap hi back in */ \
>> +       jt, \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1) /* failed: swap hi back in */
>> +
>> +#define JNE64(lo, hi, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (hi), 5, 0), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 0), /* swap in lo */ \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (lo), 2, 0), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1), /* passed: swap hi back in */ \
>> +       jt, \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1) /* failed: swap hi back in */
>> +
>> +#define JA32(value, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JSET+BPF_K, (value), 0, 1), \
>> +       jt
>> +
>> +#define JA64(lo, hi, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JSET+BPF_K, (hi), 3, 0), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 0), /* swap in lo */ \
>> +       BPF_JUMP(BPF_JMP+BPF_JSET+BPF_K, (lo), 0, 2), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1), /* passed: swap hi back in */ \
>> +       jt, \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1) /* failed: swap hi back in */
>> +
>> +#define JGE32(value, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGE+BPF_K, (value), 0, 1), \
>> +       jt
>> +
>> +#define JLT32(value, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGE+BPF_K, (value), 1, 0), \
>> +       jt
>> +
>> +/* Shortcut checking if hi > arg.hi. */
>> +#define JGE64(lo, hi, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (hi), 4, 0), \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (hi), 0, 5), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 0), /* swap in lo */ \
>> +       BPF_JUMP(BPF_JMP+BPF_JGE+BPF_K, (lo), 0, 2), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1), /* passed: swap hi back in */ \
>> +       jt, \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1) /* failed: swap hi back in */
>> +
>> +#define JLT64(lo, hi, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGE+BPF_K, (hi), 0, 4), \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (hi), 0, 5), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 0), /* swap in lo */ \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (lo), 2, 0), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1), /* passed: swap hi back in */ \
>> +       jt, \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1) /* failed: swap hi back in */
>> +
>> +#define JGT32(value, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (value), 0, 1), \
>> +       jt
>> +
>> +#define JLE32(value, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (value), 1, 0), \
>> +       jt
>> +
>> +/* Check hi > args.hi first, then do the GE checking */
>> +#define JGT64(lo, hi, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (hi), 4, 0), \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (hi), 0, 5), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 0), /* swap in lo */ \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (lo), 0, 2), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1), /* passed: swap hi back in */ \
>> +       jt, \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1) /* failed: swap hi back in */
>> +
>> +#define JLE64(lo, hi, jt) \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (hi), 6, 0), \
>> +       BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, (hi), 0, 3), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 0), /* swap in lo */ \
>> +       BPF_JUMP(BPF_JMP+BPF_JGT+BPF_K, (lo), 2, 0), \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1), /* passed: swap hi back in */ \
>> +       jt, \
>> +       BPF_STMT(BPF_LD+BPF_MEM, 1) /* failed: swap hi back in */
>> +
>> +#define LOAD_SYSCALL_NR \
>> +       BPF_STMT(BPF_LD+BPF_W+BPF_ABS, \
>> +                offsetof(struct seccomp_data, nr))
>> +
>> +#endif  /* __BPF_HELPER_H__ */
>> diff --git a/samples/seccomp/dropper.c b/samples/seccomp/dropper.c
>> new file mode 100644
>> index 0000000..c69c347
>> --- /dev/null
>> +++ b/samples/seccomp/dropper.c
>> @@ -0,0 +1,68 @@
>> +/*
>> + * Naive system call dropper built on seccomp_filter.
>> + *
>> + * Copyright (c) 2012 The Chromium OS Authors <chromium-os-dev@chromium.org>
>> + * Author: Will Drewry <wad@chromium.org>
>> + *
>> + * The code may be used by anyone for any purpose,
>> + * and can serve as a starting point for developing
>> + * applications using prctl(PR_SET_SECCOMP, 2, ...).
>> + *
>> + * When run, returns the specified errno for the specified
>> + * system call number against the given architecture.
>> + *
>> + * Run this one as root as PR_SET_NO_NEW_PRIVS is not called.
>> + */
>> +
>> +#include <errno.h>
>> +#include <linux/audit.h>
>> +#include <linux/filter.h>
>> +#include <linux/seccomp.h>
>> +#include <linux/unistd.h>
>> +#include <stdio.h>
>> +#include <stddef.h>
>> +#include <stdlib.h>
>> +#include <sys/prctl.h>
>> +#include <unistd.h>
>> +
>> +static int install_filter(int nr, int arch, int error)
>> +{
>> +       struct sock_filter filter[] = {
>> +               BPF_STMT(BPF_LD+BPF_W+BPF_ABS,
>> +                        (offsetof(struct seccomp_data, arch))),
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, arch, 0, 3),
>> +               BPF_STMT(BPF_LD+BPF_W+BPF_ABS,
>> +                        (offsetof(struct seccomp_data, nr))),
>> +               BPF_JUMP(BPF_JMP+BPF_JEQ+BPF_K, nr, 0, 1),
>> +               BPF_STMT(BPF_RET+BPF_K,
>> +                        SECCOMP_RET_ERRNO|(error & SECCOMP_RET_DATA)),
>> +               BPF_STMT(BPF_RET+BPF_K, SECCOMP_RET_ALLOW),
>> +       };
>> +       struct sock_fprog prog = {
>> +               .len = (unsigned short)(sizeof(filter)/sizeof(filter[0])),
>> +               .filter = filter,
>> +       };
>> +       if (prctl(PR_SET_SECCOMP, 2, &prog)) {
>> +               perror("prctl");
>> +               return 1;
>> +       }
>> +       return 0;
>> +}
>> +
>> +int main(int argc, char **argv)
>> +{
>> +       if (argc < 5) {
>> +               fprintf(stderr, "Usage:\n"
>> +                       "dropper <syscall_nr> <arch> <errno> <prog> [<args>]\n"
>> +                       "Hint:  AUDIT_ARCH_I386: 0x%X\n"
>> +                       "       AUDIT_ARCH_X86_64: 0x%X\n"
>> +                       "\n", AUDIT_ARCH_I386, AUDIT_ARCH_X86_64);
>> +               return 1;
>> +       }
>> +       if (install_filter(strtol(argv[1], NULL, 0), strtol(argv[2], NULL, 0),
>> +                          strtol(argv[3], NULL, 0)))
>> +               return 1;
>> +       execv(argv[4], &argv[4]);
>> +       printf("Failed to execv\n");
>> +       return 255;
>> +}
>> --
>> 1.7.5.4
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe netdev" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2012-04-18 22:37    [W:0.159 / U:0.752 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site