seccomp_unotify.2: Document SECCOMP_ADDFD_FLAG_SEND

[man-pages.git] / man2 / seccomp_unotify.2

.\" Copyright (C) 2020 Michael Kerrisk <mtk.manpages@gmail.com>
.\"
.\" %%%LICENSE_START(VERBATIM)
.\" Permission is granted to make and distribute verbatim copies of this
.\" manual provided the copyright notice and this permission notice are
.\" preserved on all copies.
.\"
.\" Permission is granted to copy and distribute modified versions of this
.\" manual under the conditions for verbatim copying, provided that the
.\" entire resulting derived work is distributed under the terms of a
.\" permission notice identical to this one.
.\"
.\" Since the Linux kernel and libraries are constantly changing, this
.\" manual page may be incorrect or out-of-date.  The author(s) assume no
.\" responsibility for errors or omissions, or for damages resulting from
.\" the use of the information contained herein.  The author(s) may not
.\" have taken the same level of care in the production of this manual,
.\" which is licensed free of charge, as they might when working
.\" professionally.
.\"
.\" Formatted or processed versions of this manual, if unaccompanied by
.\" the source, must acknowledge the copyright and authors of this work.
.\" %%%LICENSE_END
.\"
.TH SECCOMP_UNOTIFY 2 2021-06-20 "Linux" "Linux Programmer's Manual"
.SH NAME
seccomp_unotify \- Seccomp user-space notification mechanism
.SH SYNOPSIS
.nf
.B #include <linux/seccomp.h>
.B #include <linux/filter.h>
.B #include <linux/audit.h>
.PP
.BI "int seccomp(unsigned int " operation ", unsigned int " flags \
", void *" args );
.PP
.B #include <sys/ioctl.h>
.PP
.BI "int ioctl(int " fd ", SECCOMP_IOCTL_NOTIF_RECV,"
.BI "          struct seccomp_notif *" req );
.BI "int ioctl(int " fd ", SECCOMP_IOCTL_NOTIF_SEND,"
.BI "          struct seccomp_notif_resp *" resp );
.BI "int ioctl(int " fd ", SECCOMP_IOCTL_NOTIF_ID_VALID, __u64 *" id );
.BI "int ioctl(int " fd ", SECCOMP_IOCTL_NOTIF_ADDFD,"
.BI "          struct seccomp_notif_addfd *" addfd );
.fi
.SH DESCRIPTION
This page describes the user-space notification mechanism provided by the
Secure Computing (seccomp) facility.
As well as the use of the
.B SECCOMP_FILTER_FLAG_NEW_LISTENER
flag, the
.BR SECCOMP_RET_USER_NOTIF
action value, and the
.B SECCOMP_GET_NOTIF_SIZES
operation described in
.BR seccomp (2),
this mechanism involves the use of a number of related
.BR ioctl (2)
operations (described below).
.\"
.SS Overview
In conventional usage of a seccomp filter,
the decision about how to treat a system call is made by the filter itself.
By contrast, the user-space notification mechanism allows
the seccomp filter to delegate
the handling of the system call to another user-space process.
Note that this mechanism is explicitly
.B not
intended as a method implementing security policy; see NOTES.
.PP
In the discussion that follows,
the thread(s) on which the seccomp filter is installed is (are)
referred to as the
.IR target ,
and the process that is notified by the user-space notification
mechanism is referred to as the
.IR supervisor .
.PP
A suitably privileged supervisor can use the user-space notification
mechanism to perform actions on behalf of the target.
The advantage of the user-space notification mechanism is that
the supervisor will
usually be able to retrieve information about the target and the
performed system call that the seccomp filter itself cannot.
(A seccomp filter is limited in the information it can obtain and
the actions that it can perform because it
is running on a virtual machine inside the kernel.)
.PP
An overview of the steps performed by the target and the supervisor
is as follows:
.\"-------------------------------------
.IP 1. 3
The target establishes a seccomp filter in the usual manner,
but with two differences:
.RS
.IP \(bu 2
The
.BR seccomp (2)
.I flags
argument includes the flag
.BR SECCOMP_FILTER_FLAG_NEW_LISTENER .
Consequently, the return value of the (successful)
.BR seccomp (2)
call is a new "listening"
file descriptor that can be used to receive notifications.
Only one "listening" seccomp filter can be installed for a thread.
.\" FIXME
.\" Is the last sentence above correct?
.\"
.\" Kees Cook (25 Oct 2020) notes:
.\"
.\" I like this limitation, but I expect that it'll need to change in the
.\" future. Even with LSMs, we see the need for arbitrary stacking, and the
.\" idea of there being only 1 supervisor will eventually break down. Right
.\" now there is only 1 because only container managers are using this
.\" feature. But if some daemon starts using it to isolate some thread,
.\" suddenly it might break if a container manager is trying to listen to it
.\" too, etc. I expect it won't be needed soon, but I do think it'll change.
.\"
.IP \(bu
In cases where it is appropriate, the seccomp filter returns the action value
.BR SECCOMP_RET_USER_NOTIF .
This return value will trigger a notification event.
.RE
.\"-------------------------------------
.IP 2.
In order that the supervisor can obtain notifications
using the listening file descriptor,
(a duplicate of) that file descriptor must be passed from
the target to the supervisor.
One way in which this could be done is by passing the file descriptor
over a UNIX domain socket connection between the target and the supervisor
(using the
.BR SCM_RIGHTS
ancillary message type described in
.BR unix (7)).
Another way to do this is through the use of
.BR pidfd_getfd (2).
.\" Jann Horn:
.\"     Instead of using unix domain sockets to send the fd to the
.\"     parent, I think you could also use clone3() with
.\"     flags==CLONE_FILES|SIGCHLD, dup2() the seccomp fd to an fd
.\"     that was reserved in the parent, call unshare(CLONE_FILES)
.\"     in the child after setting up the seccomp fd, and wake
.\"     up the parent with something like pthread_cond_signal()?
.\"     I'm not sure whether that'd look better or worse in the
.\"     end though, so maybe just ignore this comment.
.\"-------------------------------------
.IP 3.
The supervisor will receive notification events
on the listening file descriptor.
These events are returned as structures of type
.IR seccomp_notif .
Because this structure and its size may evolve over kernel versions,
the supervisor must first determine the size of this structure
using the
.BR seccomp (2)
.B SECCOMP_GET_NOTIF_SIZES
operation, which returns a structure of type
.IR seccomp_notif_sizes .
The supervisor allocates a buffer of size
.I seccomp_notif_sizes.seccomp_notif
bytes to receive notification events.
In addition,the supervisor allocates another buffer of size
.I seccomp_notif_sizes.seccomp_notif_resp
bytes for the response (a
.I struct seccomp_notif_resp
structure)
that it will provide to the kernel (and thus the target).
.\"-------------------------------------
.IP 4.
The target then performs its workload,
which includes system calls that will be controlled by the seccomp filter.
Whenever one of these system calls causes the filter to return the
.B SECCOMP_RET_USER_NOTIF
action value, the kernel does
.I not
(yet) execute the system call;
instead, execution of the target is temporarily blocked inside
the kernel (in a sleep state that is interruptible by signals)
and a notification event is generated on the listening file descriptor.
.\"-------------------------------------
.IP 5.
The supervisor can now repeatedly monitor the
listening file descriptor for
.BR SECCOMP_RET_USER_NOTIF -triggered
events.
To do this, the supervisor uses the
.B SECCOMP_IOCTL_NOTIF_RECV
.BR ioctl (2)
operation to read information about a notification event;
this operation blocks until an event is available.
The operation returns a
.I seccomp_notif
structure containing information about the system call
that is being attempted by the target.
(As described in NOTES,
the file descriptor can also be monitored with
.BR select (2),
.BR poll (2),
or
.BR epoll (7).)
.\" FIXME
.\" Christian Brauner:
.\"
.\" Do we support O_NONBLOCK with SECCOMP_IOCTL_NOTIF_RECV and if
.\" not should we?
.\"
.\" Michael Kerrisk:
.\"
.\" A quick test suggests that O_NONBLOCK has no effect on the blocking
.\" behavior of SECCOMP_IOCTL_NOTIF_RECV.
.
.\"-------------------------------------
.IP 6.
The
.I seccomp_notif
structure returned by the
.B SECCOMP_IOCTL_NOTIF_RECV
operation includes the same information (a
.I seccomp_data
structure) that was passed to the seccomp filter.
This information allows the supervisor to discover the system call number and
the arguments for the target's system call.
In addition, the notification event contains the ID of the thread
that triggered the notification and a unique cookie value that
is used in subsequent
.B SECCOMP_IOCTL_NOTIF_ID_VALID
and
.B SECCOMP_IOCTL_NOTIF_SEND
operations.
.IP
The information in the notification can be used to discover the
values of pointer arguments for the target's system call.
(This is something that can't be done from within a seccomp filter.)
One way in which the supervisor can do this is to open the corresponding
.I /proc/[tid]/mem
file (see
.BR proc (5))
and read bytes from the location that corresponds to one of
the pointer arguments whose value is supplied in the notification event.
.\" Tycho Andersen mentioned that there are alternatives to /proc/PID/mem,
.\" such as ptrace() and /proc/PID/map_files
(The supervisor must be careful to avoid
a race condition that can occur when doing this;
see the description of the
.BR SECCOMP_IOCTL_NOTIF_ID_VALID
.BR ioctl (2)
operation below.)
In addition,
the supervisor can access other system information that is visible
in user space but which is not accessible from a seccomp filter.
.\"-------------------------------------
.IP 7.
Having obtained information as per the previous step,
the supervisor may then choose to perform an action in response
to the target's system call
(which, as noted above, is not executed when the seccomp filter returns the
.B SECCOMP_RET_USER_NOTIF
action value).
.IP
One example use case here relates to containers.
The target may be located inside a container where
it does not have sufficient capabilities to mount a filesystem
in the container's mount namespace.
However, the supervisor may be a more privileged process that
does have sufficient capabilities to perform the mount operation.
.\"-------------------------------------
.IP 8.
The supervisor then sends a response to the notification.
The information in this response is used by the kernel to construct
a return value for the target's system call and provide
a value that will be assigned to the
.I errno
variable of the target.
.IP
The response is sent using the
.B SECCOMP_IOCTL_NOTIF_SEND
.BR ioctl (2)
operation, which is used to transmit a
.I seccomp_notif_resp
structure to the kernel.
This structure includes a cookie value that the supervisor obtained in the
.I seccomp_notif
structure returned by the
.B SECCOMP_IOCTL_NOTIF_RECV
operation.
This cookie value allows the kernel to associate the response with the
target.
This structure must include the cookie value that the supervisor
obtained in the
.I seccomp_notif
structure returned by the
.B SECCOMP_IOCTL_NOTIF_RECV
operation;
the cookie allows the kernel to associate the response with the target.
.\"-------------------------------------
.IP 9.
Once the notification has been sent,
the system call in the target thread unblocks,
returning the information that was provided by the supervisor
in the notification response.
.\"-------------------------------------
.PP
As a variation on the last two steps,
the supervisor can send a response that tells the kernel that it
should execute the target thread's system call; see the discussion of
.BR SECCOMP_USER_NOTIF_FLAG_CONTINUE ,
below.
.\"
.SH IOCTL OPERATIONS
The following
.BR ioctl (2)
operations are supported by the seccomp user-space
notification file descriptor.
For each of these operations, the first (file descriptor) argument of
.BR ioctl (2)
is the listening file descriptor returned by a call to
.BR seccomp (2)
with the
.BR SECCOMP_FILTER_FLAG_NEW_LISTENER
flag.
.\"
.SS SECCOMP_IOCTL_NOTIF_RECV
The
.B SECCOMP_IOCTL_NOTIF_RECV
operation (available since Linux 5.0) is used to obtain a user-space
notification event.
If no such event is currently pending,
the operation blocks until an event occurs.
The third
.BR ioctl (2)
argument is a pointer to a structure of the following form
which contains information about the event.
This structure must be zeroed out before the call.
.PP
.in +4n
.EX
struct seccomp_notif {
    __u64  id;              /* Cookie */
    __u32  pid;             /* TID of target thread */
    __u32  flags;           /* Currently unused (0) */
    struct seccomp_data data;   /* See seccomp(2) */
};
.EE
.in
.PP
The fields in this structure are as follows:
.TP
.I id
This is a cookie for the notification.
Each such cookie is guaranteed to be unique for the corresponding
seccomp filter.
.RS
.IP \(bu 2
The cookie can be used with the
.B SECCOMP_IOCTL_NOTIF_ID_VALID
.BR ioctl (2)
operation described below.
.IP \(bu
When returning a notification response to the kernel,
the supervisor must include the cookie value in the
.IR seccomp_notif_resp
structure that is specified as the argument of the
.BR SECCOMP_IOCTL_NOTIF_SEND
operation.
.RE
.TP
.I pid
This is the thread ID of the target thread that triggered
the notification event.
.TP
.I flags
This is a bit mask of flags providing further information on the event.
In the current implementation, this field is always zero.
.TP
.I data
This is a
.I seccomp_data
structure containing information about the system call that
triggered the notification.
This is the same structure that is passed to the seccomp filter.
See
.BR seccomp (2)
for details of this structure.
.PP
On success, this operation returns 0; on failure, \-1 is returned, and
.I errno
is set to indicate the cause of the error.
This operation can fail with the following errors:
.TP
.BR EINVAL " (since Linux 5.5)"
.\" commit 2882d53c9c6f3b8311d225062522f03772cf0179
The
.I seccomp_notif
structure that was passed to the call contained nonzero fields.
.TP
.B ENOENT
The target thread was killed by a signal as the notification information
was being generated,
or the target's (blocked) system call was interrupted by a signal handler.
.\" FIXME
.\" From my experiments,
.\" it appears that if a SECCOMP_IOCTL_NOTIF_RECV is done after
.\" the target thread terminates, then the ioctl() simply
.\" blocks (rather than returning an error to indicate that the
.\" target no longer exists).
.\"
.\" I found that surprising, and it required some contortions in
.\" the example program.  It was not possible to code my SIGCHLD
.\" handler (which reaps the zombie when the worker/target
.\" terminates) to simply set a flag checked in the main
.\" handleNotifications() loop, since this created an
.\" unavoidable race where the child might terminate just after
.\" I had checked the flag, but before I blocked (forever!) in the
.\" SECCOMP_IOCTL_NOTIF_RECV operation. Instead, I had to code
.\" the signal handler to simply call _exit(2) in order to
.\" terminate the parent process (the supervisor).
.\"
.\" Is this expected behavior? It seems to me rather
.\" desirable that SECCOMP_IOCTL_NOTIF_RECV should give an error
.\" if the target has terminated.
.\"
.\" Jann posted a patch to rectify this, but there was no response
.\" (Lore link: https://bit.ly/3jvUBxk) to his question about fixing
.\" this issue. (I've tried building with the patch, but encountered
.\" an issue with the target process entering D state after a signal.)
.\"
.\" For now, this behavior is documented in BUGS.
.\"
.\" Kees Cook commented: Let's change [this] ASAP!
.\"
.SS SECCOMP_IOCTL_NOTIF_ID_VALID
The
.B SECCOMP_IOCTL_NOTIF_ID_VALID
operation (available since Linux 5.0) is used to check that a notification ID
returned by an earlier
.B SECCOMP_IOCTL_NOTIF_RECV
operation is still valid
(i.e., that the target still exists and its system call
is still blocked waiting for a response).
.PP
The third
.BR ioctl (2)
argument is a pointer to the cookie
.RI ( id )
returned by the
.B SECCOMP_IOCTL_NOTIF_RECV
operation.
.PP
This operation is necessary to avoid race conditions that can occur when the
.I pid
returned by the
.B SECCOMP_IOCTL_NOTIF_RECV
operation terminates, and that process ID is reused by another process.
An example of this kind of race is the following
.IP 1. 3
A notification is generated on the listening file descriptor.
The returned
.I seccomp_notif
contains the TID of the target thread (in the
.I pid
field of the structure).
.IP 2.
The target terminates.
.IP 3.
Another thread or process is created on the system that by chance reuses the
TID that was freed when the target terminated.
.IP 4.
The supervisor
.BR open (2)s
the
.IR /proc/[tid]/mem
file for the TID obtained in step 1, with the intention of (say)
inspecting the memory location(s) that containing the argument(s) of
the system call that triggered the notification in step 1.
.PP
In the above scenario, the risk is that the supervisor may try
to access the memory of a process other than the target.
This race can be avoided by following the call to
.BR open (2)
with a
.B SECCOMP_IOCTL_NOTIF_ID_VALID
operation to verify that the process that generated the notification
is still alive.
(Note that if the target terminates after the latter step,
a subsequent
.BR read (2)
from the file descriptor may return 0, indicating end of file.)
.\" Jann Horn:
.\"     the PID can be reused, but the /proc/$pid directory is
.\"     internally not associated with the numeric PID, but,
.\"     conceptually speaking, with a specific incarnation of the
.\"     PID, or something like that.  (Actually, it is associated
.\"     with the "struct pid", which is not reused, instead of the
.\"     numeric PID.
.PP
See NOTES for a discussion of other cases where
.B SECCOMP_IOCTL_NOTIF_ID_VALID
checks must be performed.
.PP
On success (i.e., the notification ID is still valid),
this operation returns 0.
On failure (i.e., the notification ID is no longer valid),
\-1 is returned, and
.I errno
is set to
.BR ENOENT .
.\"
.SS SECCOMP_IOCTL_NOTIF_SEND
The
.B SECCOMP_IOCTL_NOTIF_SEND
operation (available since Linux 5.0)
is used to send a notification response back to the kernel.
The third
.BR ioctl (2)
argument of this structure is a pointer to a structure of the following form:
.PP
.in +4n
.EX
struct seccomp_notif_resp {
    __u64 id;           /* Cookie value */
    __s64 val;          /* Success return value */
    __s32 error;        /* 0 (success) or negative error number */
    __u32 flags;        /* See below */
};
.EE
.in
.PP
The fields of this structure are as follows:
.TP
.I id
This is the cookie value that was obtained using the
.B SECCOMP_IOCTL_NOTIF_RECV
operation.
This cookie value allows the kernel to correctly associate this response
with the system call that triggered the user-space notification.
.TP
.I val
This is the value that will be used for a spoofed
success return for the target's system call; see below.
.TP
.I error
This is the value that will be used as the error number
.RI ( errno )
for a spoofed error return for the target's system call; see below.
.TP
.I flags
This is a bit mask that includes zero or more of the following flags:
.RS
.TP
.BR SECCOMP_USER_NOTIF_FLAG_CONTINUE " (since Linux 5.5)"
Tell the kernel to execute the target's system call.
.\" commit fb3c5386b382d4097476ce9647260fc89b34afdb
.RE
.PP
Two kinds of response are possible:
.IP \(bu 2
A response to the kernel telling it to execute the
target's system call.
In this case, the
.I flags
field includes
.B SECCOMP_USER_NOTIF_FLAG_CONTINUE
and the
.I error
and
.I val
fields must be zero.
.IP
This kind of response can be useful in cases where the supervisor needs
to do deeper analysis of the target's system call than is possible
from a seccomp filter (e.g., examining the values of pointer arguments),
and, having decided that the system call does not require emulation
by the supervisor, the supervisor wants the system call to
be executed normally in the target.
.IP
The
.B SECCOMP_USER_NOTIF_FLAG_CONTINUE
flag should be used with caution; see NOTES.
.IP \(bu
A spoofed return value for the target's system call.
In this case, the kernel does not execute the target's system call,
instead causing the system call to return a spoofed value as specified by
fields of the
.I seccomp_notif_resp
structure.
The supervisor should set the fields of this structure as follows:
.RS
.IP + 3
.I flags
does not contain
.BR SECCOMP_USER_NOTIF_FLAG_CONTINUE .
.IP +
.I error
is set either to 0 for a spoofed "success" return or to a negative
error number for a spoofed "failure" return.
In the former case, the kernel causes the target's system call
to return the value specified in the
.I val
field.
In the latter case, the kernel causes the target's system call
to return \-1, and
.I errno
is assigned the negated
.I error
value.
.IP +
.I val
is set to a value that will be used as the return value for a spoofed
"success" return for the target's system call.
The value in this field is ignored if the
.I error
field contains a nonzero value.
.\" FIXME
.\" Kees Cook suggested:
.\"
.\" Strictly speaking, this is architecture specific, but
.\" all architectures do it this way. Should seccomp enforce
.\" val == 0 when err != 0 ?
.\"
.\" Christian Brauner
.\"
.\" Feels like it should, at least for the SEND ioctl where we already
.\" verify that val and err are both 0 when CONTINUE is specified (as you
.\" pointed out correctly above).
.RE
.PP
On success, this operation returns 0; on failure, \-1 is returned, and
.I errno
is set to indicate the cause of the error.
This operation can fail with the following errors:
.TP
.B EINPROGRESS
A response to this notification has already been sent.
.TP
.B EINVAL
An invalid value was specified in the
.I flags field.
.TP
.B
.B EINVAL
The
.I flags
field contained
.BR SECCOMP_USER_NOTIF_FLAG_CONTINUE ,
and the
.I error
or
.I val
field was not zero.
.TP
.B ENOENT
The blocked system call in the target
has been interrupted by a signal handler
or the target has terminated.
.\" Jann Horn notes:
.\"     you could also get this [ENOENT] if a response has already
.\"     been sent, instead of EINPROGRESS - the only difference is
.\"     whether the target thread has picked up the response yet
.\"
.SS SECCOMP_IOCTL_NOTIF_ADDFD
The
.B SECCOMP_IOCTL_NOTIF_ADDFD
operation (available since Linux 5.9)
allows the supervisor to install a file descriptor
into the target's file descriptor table.
Much like the use of
.BR SCM_RIGHTS
messages described in
.BR unix (7),
this operation is semantically equivalent to duplicating
a file descriptor from the supervisor's file descriptor table
into the target's file descriptor table.
.PP
The
.BR SECCOMP_IOCTL_NOTIF_ADDFD
operation permits the supervisor to emulate a target system call (such as
.BR socket (2)
or
.BR openat (2))
that generates a file descriptor.
The supervisor can perform the system call that generates
the file descriptor (and associated open file description)
and then use this operation to allocate
a file descriptor that refers to the same open file description in the target.
(For an explanation of open file descriptions, see
.BR open (2).)
.PP
Once this operation has been performed,
the supervisor can close its copy of the file descriptor.
.PP
In the target,
the received file descriptor is subject to the same
Linux Security Module (LSM) checks as are applied to a file descriptor
that is received in an
.BR SCM_RIGHTS
ancillary message.
If the file descriptor refers to a socket,
it inherits the cgroup version 1 network controller settings
.RI ( classid
and
.IR netprioidx )
of the target.
.PP
The third
.BR ioctl (2)
argument is a pointer to a structure of the following form:
.PP
.in +4n
.EX
struct seccomp_notif_addfd {
    __u64 id;           /* Cookie value */
    __u32 flags;        /* Flags */
    __u32 srcfd;        /* Local file descriptor number */
    __u32 newfd;        /* 0 or desired file descriptor
                           number in target */
    __u32 newfd_flags;  /* Flags to set on target file
                           descriptor */
};
.EE
.in
.PP
The fields in this structure are as follows:
.TP
.I id
This field should be set to the notification ID
(cookie value) that was obtained via
.BR SECCOMP_IOCTL_NOTIF_RECV .
.TP
.I flags
This field is a bit mask of flags that modify the behavior of the operation.
Currently, only one flag is supported:
.RS
.TP
.BR SECCOMP_ADDFD_FLAG_SETFD
When allocating the file descriptor in the target,
use the file descriptor number specified in the
.I newfd
field.
.TP
.BR SECCOMP_ADDFD_FLAG_SEND
Available since Linux 5.14, combines the
.B SECCOMP_IOCTL_NOTIF_ADDFD
ioctl with
.B SECCOMP_IOCTL_NOTIF_SEND
into an atomic operation. On successful invocation, the target process's
errno will be 0 and the return value will be the file descriptor number that was
installed in the target. If allocating the file descriptor in the tatget fails,
the target's syscall continues to be blocked until a successful response is
sent.
.RE
.TP
.I srcfd
This field should be set to the number of the file descriptor
in the supervisor that is to be duplicated.
.TP
.I newfd
This field determines which file descriptor number is allocated in the target.
If the
.BR SECCOMP_ADDFD_FLAG_SETFD
flag is set,
then this field specifies which file descriptor number should be allocated.
If this file descriptor number is already open in the target,
it is atomically closed and reused.
If the descriptor duplication fails due to an LSM check, or if
.I srcfd
is not a valid file descriptor,
the file descriptor
.I newfd
will not be closed in the target process.
.IP
If the
.BR SECCOMP_ADDFD_FLAG_SETFD
flag it not set, then this field must be 0,
and the kernel allocates the lowest unused file descriptor number
in the target.
.TP
.I newfd_flags
This field is a bit mask specifying flags that should be set on
the file descriptor that is received in the target process.
Currently, only the following flag is implemented:
.RS
.TP
.B O_CLOEXEC
Set the close-on-exec flag on the received file descriptor.
.RE
.PP
On success, this
.BR ioctl (2)
call returns the number of the file descriptor that was allocated
in the target.
Assuming that the emulated system call is one that returns
a file descriptor as its function result (e.g.,
.BR socket (2)),
this value can be used as the return value
.RI ( resp.val )
that is supplied in the response that is subsequently sent with the
.BR SECCOMP_IOCTL_NOTIF_SEND
operation.
.PP
On error, \-1 is returned and
.I errno
is set to indicate the cause of the error.
.PP
This operation can fail with the following errors:
.TP
.B EBADF
Allocating the file descriptor in the target would cause the target's
.BR RLIMIT_NOFILE
limit to be exceeded (see
.BR getrlimit (2)).
.TP
.B EBUSY
If the flag
.B SECCOMP_IOCTL_NOTIF_SEND
is used, this means the operation can't proceed until other
.B SECCOMP_IOCTL_NOTIF_ADDFD
requests are processed.
.TP
.B EINPROGRESS
The user-space notification specified in the
.I id
field exists but has not yet been fetched (by a
.BR SECCOMP_IOCTL_NOTIF_RECV )
or has already been responded to (by a
.BR SECCOMP_IOCTL_NOTIF_SEND ).
.TP
.B EINVAL
An invalid flag was specified in the
.I flags
or
.I newfd_flags
field, or the
.I newfd
field is nonzero and the
.B SECCOMP_ADDFD_FLAG_SETFD
flag was not specified in the
.I flags
field.
.TP
.B EMFILE
The file descriptor number specified in
.I newfd
exceeds the limit specified in
.IR /proc/sys/fs/nr_open .
.TP
.B ENOENT
The blocked system call in the target
has been interrupted by a signal handler
or the target has terminated.
.PP
Here is some sample code (with error handling omitted) that uses the
.B SECCOMP_ADDFD_FLAG_SETFD
operation (here, to emulate a call to
.BR openat (2)):
.PP
.EX
.in +4n
int fd, removeFd;

fd = openat(req->data.args[0], path, req->data.args[2],
                req->data.args[3]);

struct seccomp_notif_addfd addfd;
addfd.id = req->id; /* Cookie from SECCOMP_IOCTL_NOTIF_RECV */
addfd.srcfd = fd;
addfd.newfd = 0;
addfd.flags = 0;
addfd.newfd_flags = O_CLOEXEC;

targetFd = ioctl(notifyFd, SECCOMP_IOCTL_NOTIF_ADDFD, &addfd);

close(fd);          /* No longer needed in supervisor */

struct seccomp_notif_resp *resp;
    /* Code to allocate 'resp' omitted */
resp->id = req->id;
resp->error = 0;        /* "Success" */
resp->val = targetFd;
resp->flags = 0;
ioctl(notifyFd, SECCOMP_IOCTL_NOTIF_SEND, resp);
.in
.EE
.SH NOTES
One example use case for the user-space notification
mechanism is to allow a container manager
(a process which is typically running with more privilege than
the processes inside the container)
to mount block devices or create device nodes for the container.
The mount use case provides an example of where the
.BR SECCOMP_USER_NOTIF_FLAG_CONTINUE
.BR ioctl (2)
operation is useful.
Upon receiving a notification for the
.BR mount (2)
system call, the container manager (the "supervisor") can distinguish
a request to mount a block filesystem
(which would not be possible for a "target" process inside the container)
and mount that file system.
If, on the other hand, the container manager detects that the operation
could be performed by the process inside the container
(e.g., a mount of a
.BR tmpfs (5)
filesystem), it can notify the kernel that the target process's
.BR mount (2)
system call can continue.
.\"
.SS select()/poll()/epoll semantics
The file descriptor returned when
.BR seccomp (2)
is employed with the
.B SECCOMP_FILTER_FLAG_NEW_LISTENER
flag can be monitored using
.BR poll (2),
.BR epoll (7),
and
.BR select (2).
These interfaces indicate that the file descriptor is ready as follows:
.IP \(bu 2
When a notification is pending,
these interfaces indicate that the file descriptor is readable.
Following such an indication, a subsequent
.B SECCOMP_IOCTL_NOTIF_RECV
.BR ioctl (2)
will not block, returning either information about a notification
or else failing with the error
.B EINTR
if the target has been killed by a signal or its system call
has been interrupted by a signal handler.
.IP \(bu
After the notification has been received (i.e., by the
.B SECCOMP_IOCTL_NOTIF_RECV
.BR ioctl (2)
operation), these interfaces indicate that the file descriptor is writable,
meaning that a notification response can be sent using the
.B SECCOMP_IOCTL_NOTIF_SEND
.BR ioctl (2)
operation.
.IP \(bu
After the last thread using the filter has terminated and been reaped using
.BR waitpid (2)
(or similar),
the file descriptor indicates an end-of-file condition (readable in
.BR select (2);
.BR POLLHUP / EPOLLHUP
in
.BR poll (2)/
.BR epoll_wait (2)).
.SS Design goals; use of SECCOMP_USER_NOTIF_FLAG_CONTINUE
The intent of the user-space notification feature is
to allow system calls to be performed on behalf of the target.
The target's system call should either be handled by the supervisor or
allowed to continue normally in the kernel (where standard security
policies will be applied).
.PP
.BR "Note well" :
this mechanism must not be used to make security policy decisions
about the system call,
which would be inherently race-prone for reasons described next.
.PP
The
.B SECCOMP_USER_NOTIF_FLAG_CONTINUE
flag must be used with caution.
If set by the supervisor, the target's system call will continue.
However, there is a time-of-check, time-of-use race here,
since an attacker could exploit the interval of time where the target is
blocked waiting on the "continue" response to do things such as
rewriting the system call arguments.
.PP
Note furthermore that a user-space notifier can be bypassed if
the existing filters allow the use of
.BR seccomp (2)
or
.BR prctl (2)
to install a filter that returns an action value with a higher precedence than
.B SECCOMP_RET_USER_NOTIF
(see
.BR seccomp (2)).
.PP
It should thus be absolutely clear that the
seccomp user-space notification mechanism
.B can not
be used to implement a security policy!
It should only ever be used in scenarios where a more privileged process
supervises the system calls of a lesser privileged target to
get around kernel-enforced security restrictions when
the supervisor deems this safe.
In other words,
in order to continue a system call, the supervisor should be sure that
another security mechanism or the kernel itself will sufficiently block
the system call if its arguments are rewritten to something unsafe.
.\"
.SS Caveats regarding the use of /proc/[tid]/mem
The discussion above noted the need to use the
.BR SECCOMP_IOCTL_NOTIF_ID_VALID
.BR ioctl (2)
when opening the
.IR /proc/[tid]/mem
file of the target
to avoid the possibility of accessing the memory of the wrong process
in the event that the target terminates and its ID
is recycled by another (unrelated) thread.
However, the use of this
.BR ioctl (2)
operation is also necessary in other situations,
as explained in the following paragraphs.
.PP
Consider the following scenario, where the supervisor
tries to read the pathname argument of a target's blocked
.BR mount (2)
system call:
.IP \(bu 2
From one of its functions
.RI ( func() ),
the target calls
.BR mount (2),
which triggers a user-space notification and causes the target to block.
.IP \(bu
The supervisor receives the notification, opens
.IR /proc/[tid]/mem ,
and (successfully) performs the
.BR SECCOMP_IOCTL_NOTIF_ID_VALID
check.
.IP \(bu
The target receives a signal, which causes the
.BR mount (2)
to abort.
.IP \(bu
The signal handler executes in the target, and returns.
.IP \(bu
Upon return from the handler, the execution of
.I func()
resumes, and it returns (and perhaps other functions are called,
overwriting the memory that had been used for the stack frame of
.IR func() ).
.IP \(bu
Using the address provided in the notification information,
the supervisor reads from the target's memory location that used to
contain the pathname.
.IP \(bu
The supervisor now calls
.BR mount (2)
with some arbitrary bytes obtained in the previous step.
.PP
The conclusion from the above scenario is this:
since the target's blocked system call may be interrupted by a signal handler,
the supervisor must be written to expect that the
target may abandon its system call at
.B any
time;
in such an event, any information that the supervisor obtained from
the target's memory must be considered invalid.
.PP
To prevent such scenarios,
every read from the target's memory must be separated from use of
the bytes so obtained by a
.BR SECCOMP_IOCTL_NOTIF_ID_VALID
check.
In the above example, the check would be placed between the two final steps.
An example of such a check is shown in EXAMPLES.
.PP
Following on from the above, it should be clear that
a write by the supervisor into the target's memory can
.B never
be considered safe.
.\"
.SS Caveats regarding blocking system calls
Suppose that the target performs a blocking system call (e.g.,
.BR accept (2))
that the supervisor should handle.
The supervisor might then in turn execute the same blocking system call.
.PP
In this scenario,
it is important to note that if the target's system call is now
interrupted by a signal, the supervisor is
.I not
informed of this.
If the supervisor does not take suitable steps to
actively discover that the target's system call has been canceled,
various difficulties can occur.
Taking the example of
.BR accept (2),
the supervisor might remain blocked in its
.BR accept (2)
holding a port number that the target
(which, after the interruption by the signal handler,
perhaps closed  its listening socket) might expect to be able to reuse in a
.BR bind (2)
call.
.PP
Therefore, when the supervisor wishes to emulate a blocking system call,
it must do so in such a way that it gets informed if the target's
system call is interrupted by a signal handler.
For example, if the supervisor itself executes the same
blocking system call, then it could employ a separate thread
that uses the
.B SECCOMP_IOCTL_NOTIF_ID_VALID
operation to check if the target is still blocked in its system call.
Alternatively, in the
.BR accept (2)
example, the supervisor might use
.BR poll (2)
to monitor both the notification file descriptor
(so as as to discover when the target's
.BR accept (2)
call has been interrupted) and the listening file descriptor
(so as to know when a connection is available).
.PP
If the target's system call is interrupted,
the supervisor must take care to release resources (e.g., file descriptors)
that it acquired on behalf of the target.
.\"
.SS Interaction with SA_RESTART signal handlers
Consider the following scenario:
.IP \(bu 2
The target process has used
.BR sigaction (2)
to install a signal handler with the
.B SA_RESTART
flag.
.IP \(bu
The target has made a system call that triggered a seccomp
user-space notification and the target is currently blocked
until the supervisor sends a notification response.
.IP \(bu
A signal is delivered to the target and the signal handler is executed.
.IP \(bu
When (if) the supervisor attempts to send a notification response, the
.B SECCOMP_IOCTL_NOTIF_SEND
.BR ioctl (2))
operation will fail with the
.BR ENOENT
error.
.PP
In this scenario, the kernel will restart the target's system call.
Consequently, the supervisor will receive another user-space notification.
Thus, depending on how many times the blocked system call
is interrupted by a signal handler,
the supervisor may receive multiple notifications for
the same instance of a system call in the target.
.PP
One oddity is that system call restarting as described in this scenario
will occur even for the blocking system calls listed in
.BR signal (7)
that would
.B never
normally be restarted by the
.BR SA_RESTART
flag.
.PP
Furthermore, if the supervisor response is a file descriptor
added with
.B SECCOMP_IOCTL_NOTIF_ADDFD,
then the flag
.B SECCOMP_ADDFD_FLAG_SEND
can be used to atomically add the file descriptor and return that value,
making sure no file descriptors are inadvertently leaked into the target.
.\" FIXME
.\" About the above, Kees Cook commented:
.\"
.\" Does this need fixing? I imagine the correct behavior for this case
.\" would be a response to _SEND of EINPROGRESS and the target would see
.\" EINTR normally?
.\"
.\" I mean, it's not like seccomp doesn't already expose weirdness with
.\" syscall restarts. Not even arm64 compat agrees[3] with arm32 in this
.\" regard. :(
.
.\" FIXME
.\" Michael Kerrisk:
.\" I wonder about the effect of this oddity for system calls that
.\" are normally nonrestartable because they have timeouts. My
.\" understanding is that the kernel doesn't restart those system
.\" calls because it's impossible for the kernel to restart the call
.\" with the right timeout value. I wonder what happens when those
.\" system calls are restarted in the scenario we're discussing.)
.SH BUGS
If a
.BR SECCOMP_IOCTL_NOTIF_RECV
.BR ioctl (2)
operation
.\" or a poll/epoll/select
is performed after the target terminates, then the
.BR ioctl (2)
call simply blocks (rather than returning an error to indicate that the
target no longer exists).
.\" FIXME
.\" Comment from Kees Cook:
.\"
.\" I want this fixed. It caused me no end of pain when building the
.\" selftests, and ended up spawning my implementing a global test timeout
.\" in kselftest. :P Before the usage counter refactor, there was no sane
.\" way to deal with this, but now I think we're close.
.\"
.SH EXAMPLES
The (somewhat contrived) program shown below demonstrates the use of
the interfaces described in this page.
The program creates a child process that serves as the "target" process.
The child process installs a seccomp filter that returns the
.B SECCOMP_RET_USER_NOTIF
action value if a call is made to
.BR mkdir (2).
The child process then calls
.BR mkdir (2)
once for each of the supplied command-line arguments,
and reports the result returned by the call.
After processing all arguments, the child process terminates.
.PP
The parent process acts as the supervisor, listening for the notifications
that are generated when the target process calls
.BR mkdir (2).
When such a notification occurs,
the supervisor examines the memory of the target process (using
.IR /proc/[pid]/mem )
to discover the pathname argument that was supplied to the
.BR mkdir (2)
call, and performs one of the following actions:
.IP \(bu 2
If the pathname begins with the prefix "/tmp/",
then the supervisor attempts to create the specified directory,
and then spoofs a return for the target process based on the return
value of the supervisor's
.BR mkdir (2)
call.
In the event that that call succeeds,
the spoofed success return value is the length of the pathname.
.IP \(bu
If the pathname begins with "./" (i.e., it is a relative pathname),
the supervisor sends a
.B SECCOMP_USER_NOTIF_FLAG_CONTINUE
response to the kernel to say that the kernel should execute
the target process's
.BR mkdir (2)
call.
.IP \(bu
If the pathname begins with some other prefix,
the supervisor spoofs an error return for the target process,
so that the target process's
.BR mkdir (2)
call appears to fail with the error
.BR EOPNOTSUPP
("Operation not supported").
Additionally, if the specified pathname is exactly "/bye",
then the supervisor terminates.
.PP
This program can be used to demonstrate various aspects of the
behavior of the seccomp user-space notification mechanism.
To help aid such demonstrations,
the program logs various messages to show the operation
of the target process (lines prefixed "T:") and the supervisor
(indented lines prefixed "S:").
.PP
In the following example, the target attempts to create the directory
.IR /tmp/x .
Upon receiving the notification, the supervisor creates the directory on the
target's behalf,
and spoofs a success return to be received by the target process's
.BR mkdir (2)
call.
.PP
.in +4n
.EX
$ \fB./seccomp_unotify /tmp/x\fP
T: PID = 23168

T: about to mkdir("/tmp/x")
        S: got notification (ID 0x17445c4a0f4e0e3c) for PID 23168
        S: executing: mkdir("/tmp/x", 0700)
        S: success! spoofed return = 6
        S: sending response (flags = 0; val = 6; error = 0)
T: SUCCESS: mkdir(2) returned 6

T: terminating
        S: target has terminated; bye
.EE
.in
.PP
In the above output, note that the spoofed return value seen by the target
process is 6 (the length of the pathname
.IR /tmp/x ),
whereas a normal
.BR mkdir (2)
call returns 0 on success.
.PP
In the next example, the target attempts to create a directory using the
relative pathname
.IR ./sub .
Since this pathname starts with "./",
the supervisor sends a
.B SECCOMP_USER_NOTIF_FLAG_CONTINUE
response to the kernel,
and the kernel then (successfully) executes the target process's
.BR mkdir (2)
call.
.PP
.in +4n
.EX
$ \fB./seccomp_unotify ./sub\fP
T: PID = 23204

T: about to mkdir("./sub")
        S: got notification (ID 0xddb16abe25b4c12) for PID 23204
        S: target can execute system call
        S: sending response (flags = 0x1; val = 0; error = 0)
T: SUCCESS: mkdir(2) returned 0

T: terminating
        S: target has terminated; bye
.EE
.in
.PP
If the target process attempts to create a directory with
a pathname that doesn't start with "." and doesn't begin with the prefix
"/tmp/", then the supervisor spoofs an error return
.RB ( EOPNOTSUPP ,
"Operation not  supported")
for the target's
.BR mkdir (2)
call (which is not executed):
.PP
.in +4n
.EX
$ \fB./seccomp_unotify /xxx\fP
T: PID = 23178

T: about to mkdir("/xxx")
        S: got notification (ID 0xe7dc095d1c524e80) for PID 23178
        S: spoofing error response (Operation not supported)
        S: sending response (flags = 0; val = 0; error = \-95)
T: ERROR: mkdir(2): Operation not supported

T: terminating
        S: target has terminated; bye
.EE
.in
.PP
In the next example,
the target process attempts to create a directory with the pathname
.BR /tmp/nosuchdir/b .
Upon receiving the notification,
the supervisor attempts to create that directory, but the
.BR mkdir (2)
call fails because the directory
.BR /tmp/nosuchdir
does not exist.
Consequently, the supervisor spoofs an error return that passes the error
that it received back to the target process's
.BR mkdir (2)
call.
.PP
.in +4n
.EX
$ \fB./seccomp_unotify /tmp/nosuchdir/b\fP
T: PID = 23199

T: about to mkdir("/tmp/nosuchdir/b")
        S: got notification (ID 0x8744454293506046) for PID 23199
        S: executing: mkdir("/tmp/nosuchdir/b", 0700)
        S: failure! (errno = 2; No such file or directory)
        S: sending response (flags = 0; val = 0; error = \-2)
T: ERROR: mkdir(2): No such file or directory

T: terminating
        S: target has terminated; bye
.EE
.in
.PP
If the supervisor receives a notification and sees that the
argument of the target's
.BR mkdir (2)
is the string "/bye", then (as well as spoofing an
.B EOPNOTSUPP
error), the supervisor terminates.
If the target process subsequently executes another
.BR mkdir (2)
that triggers its seccomp filter to return the
.B SECCOMP_RET_USER_NOTIF
action value, then the kernel causes the target process's system call to
fail with the error
.B ENOSYS
("Function not implemented").
This is demonstrated by the following example:
.PP
.in +4n
.EX
$ \fB./seccomp_unotify /bye /tmp/y\fP
T: PID = 23185

T: about to mkdir("/bye")
        S: got notification (ID 0xa81236b1d2f7b0f4) for PID 23185
        S: spoofing error response (Operation not supported)
        S: sending response (flags = 0; val = 0; error = \-95)
        S: terminating **********
T: ERROR: mkdir(2): Operation not supported

T: about to mkdir("/tmp/y")
T: ERROR: mkdir(2): Function not implemented

T: terminating
.EE
.in
.\"
.SS Program source
.EX
#define _GNU_SOURCE
#include <errno.h>
#include <fcntl.h>
#include <limits.h>
#include <linux/audit.h>
#include <linux/filter.h>
#include <linux/seccomp.h>
#include <signal.h>
#include <stdbool.h>
#include <stddef.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <sys/socket.h>
#include <sys/ioctl.h>
#include <sys/prctl.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <sys/un.h>
#include <sys/syscall.h>
#include <unistd.h>

#define errExit(msg)    do { perror(msg); exit(EXIT_FAILURE); \e
                        } while (0)

/* Send the file descriptor \(aqfd\(aq over the connected UNIX domain socket
   \(aqsockfd\(aq. Returns 0 on success, or \-1 on error. */

static int
sendfd(int sockfd, int fd)
{
    struct msghdr msgh;
    struct iovec iov;
    int data;
    struct cmsghdr *cmsgp;

    /* Allocate a char array of suitable size to hold the ancillary data.
       However, since this buffer is in reality a \(aqstruct cmsghdr\(aq, use a
       union to ensure that it is suitably aligned. */
    union {
        char   buf[CMSG_SPACE(sizeof(int))];
                        /* Space large enough to hold an \(aqint\(aq */
        struct cmsghdr align;
    } controlMsg;

    /* The \(aqmsg_name\(aq field can be used to specify the address of the
       destination socket when sending a datagram. However, we do not
       need to use this field because \(aqsockfd\(aq is a connected socket. */

    msgh.msg_name = NULL;
    msgh.msg_namelen = 0;

    /* On Linux, we must transmit at least one byte of real data in
       order to send ancillary data. We transmit an arbitrary integer
       whose value is ignored by recvfd(). */

    msgh.msg_iov = &iov;
    msgh.msg_iovlen = 1;
    iov.iov_base = &data;
    iov.iov_len = sizeof(int);
    data = 12345;

    /* Set \(aqmsghdr\(aq fields that describe ancillary data */

    msgh.msg_control = controlMsg.buf;
    msgh.msg_controllen = sizeof(controlMsg.buf);

    /* Set up ancillary data describing file descriptor to send */

    cmsgp = CMSG_FIRSTHDR(&msgh);
    cmsgp\->cmsg_level = SOL_SOCKET;
    cmsgp\->cmsg_type = SCM_RIGHTS;
    cmsgp\->cmsg_len = CMSG_LEN(sizeof(int));
    memcpy(CMSG_DATA(cmsgp), &fd, sizeof(int));

    /* Send real plus ancillary data */

    if (sendmsg(sockfd, &msgh, 0) == \-1)
        return \-1;

    return 0;
}

/* Receive a file descriptor on a connected UNIX domain socket. Returns
   the received file descriptor on success, or \-1 on error. */

static int
recvfd(int sockfd)
{
    struct msghdr msgh;
    struct iovec iov;
    int data, fd;
    ssize_t nr;

    /* Allocate a char buffer for the ancillary data. See the comments
       in sendfd() */
    union {
        char   buf[CMSG_SPACE(sizeof(int))];
        struct cmsghdr align;
    } controlMsg;
    struct cmsghdr *cmsgp;

    /* The \(aqmsg_name\(aq field can be used to obtain the address of the
       sending socket. However, we do not need this information. */

    msgh.msg_name = NULL;
    msgh.msg_namelen = 0;

    /* Specify buffer for receiving real data */

    msgh.msg_iov = &iov;
    msgh.msg_iovlen = 1;
    iov.iov_base = &data;       /* Real data is an \(aqint\(aq */
    iov.iov_len = sizeof(int);

    /* Set \(aqmsghdr\(aq fields that describe ancillary data */

    msgh.msg_control = controlMsg.buf;
    msgh.msg_controllen = sizeof(controlMsg.buf);

    /* Receive real plus ancillary data; real data is ignored */

    nr = recvmsg(sockfd, &msgh, 0);
    if (nr == \-1)
        return \-1;

    cmsgp = CMSG_FIRSTHDR(&msgh);

    /* Check the validity of the \(aqcmsghdr\(aq */

    if (cmsgp == NULL ||
            cmsgp\->cmsg_len != CMSG_LEN(sizeof(int)) ||
            cmsgp\->cmsg_level != SOL_SOCKET ||
            cmsgp\->cmsg_type != SCM_RIGHTS) {
        errno = EINVAL;
        return \-1;
    }

    /* Return the received file descriptor to our caller */

    memcpy(&fd, CMSG_DATA(cmsgp), sizeof(int));
    return fd;
}

static void
sigchldHandler(int sig)
{
    char msg[] = "\etS: target has terminated; bye\en";

    write(STDOUT_FILENO, msg, sizeof(msg) - 1);
    _exit(EXIT_SUCCESS);
}

static int
seccomp(unsigned int operation, unsigned int flags, void *args)
{
    return syscall(__NR_seccomp, operation, flags, args);
}

/* The following is the x86\-64\-specific BPF boilerplate code for checking
   that the BPF program is running on the right architecture + ABI. At
   completion of these instructions, the accumulator contains the system
   call number. */

/* For the x32 ABI, all system call numbers have bit 30 set */

#define X32_SYSCALL_BIT         0x40000000

#define X86_64_CHECK_ARCH_AND_LOAD_SYSCALL_NR \e
        BPF_STMT(BPF_LD | BPF_W | BPF_ABS, \e
                (offsetof(struct seccomp_data, arch))), \e
        BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, AUDIT_ARCH_X86_64, 0, 2), \e
        BPF_STMT(BPF_LD | BPF_W | BPF_ABS, \e
                 (offsetof(struct seccomp_data, nr))), \e
        BPF_JUMP(BPF_JMP | BPF_JGE | BPF_K, X32_SYSCALL_BIT, 0, 1), \e
        BPF_STMT(BPF_RET | BPF_K, SECCOMP_RET_KILL_PROCESS)

/* installNotifyFilter() installs a seccomp filter that generates
   user\-space notifications (SECCOMP_RET_USER_NOTIF) when the process
   calls mkdir(2); the filter allows all other system calls.

   The function return value is a file descriptor from which the
   user\-space notifications can be fetched. */

static int
installNotifyFilter(void)
{
    struct sock_filter filter[] = {
        X86_64_CHECK_ARCH_AND_LOAD_SYSCALL_NR,

        /* mkdir() triggers notification to user\-space supervisor */

        BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, __NR_mkdir, 0, 1),
        BPF_STMT(BPF_RET + BPF_K, SECCOMP_RET_USER_NOTIF),

        /* Every other system call is allowed */

        BPF_STMT(BPF_RET | BPF_K, SECCOMP_RET_ALLOW),
    };

    struct sock_fprog prog = {
        .len = sizeof(filter) / sizeof(filter[0]),
        .filter = filter,
    };

    /* Install the filter with the SECCOMP_FILTER_FLAG_NEW_LISTENER flag;
       as a result, seccomp() returns a notification file descriptor. */

    int notifyFd = seccomp(SECCOMP_SET_MODE_FILTER,
                           SECCOMP_FILTER_FLAG_NEW_LISTENER, &prog);
    if (notifyFd == \-1)
        errExit("seccomp\-install\-notify\-filter");

    return notifyFd;
}

/* Close a pair of sockets created by socketpair() */

static void
closeSocketPair(int sockPair[2])
{
    if (close(sockPair[0]) == \-1)
        errExit("closeSocketPair\-close\-0");
    if (close(sockPair[1]) == \-1)
        errExit("closeSocketPair\-close\-1");
}

/* Implementation of the target process; create a child process that:

   (1) installs a seccomp filter with the
       SECCOMP_FILTER_FLAG_NEW_LISTENER flag;
   (2) writes the seccomp notification file descriptor returned from
       the previous step onto the UNIX domain socket, \(aqsockPair[0]\(aq;
   (3) calls mkdir(2) for each element of \(aqargv\(aq.

   The function return value in the parent is the PID of the child
   process; the child does not return from this function. */

static pid_t
targetProcess(int sockPair[2], char *argv[])
{
    pid_t targetPid = fork();
    if (targetPid == \-1)
        errExit("fork");

    if (targetPid > 0)          /* In parent, return PID of child */
        return targetPid;

    /* Child falls through to here */

    printf("T: PID = %ld\en", (long) getpid());

    /* Install seccomp filter(s) */

    if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0))
        errExit("prctl");

    int notifyFd = installNotifyFilter();

    /* Pass the notification file descriptor to the tracing process over
       a UNIX domain socket */

    if (sendfd(sockPair[0], notifyFd) == \-1)
        errExit("sendfd");

    /* Notification and socket FDs are no longer needed in target */

    if (close(notifyFd) == \-1)
        errExit("close\-target\-notify\-fd");

    closeSocketPair(sockPair);

    /* Perform a mkdir() call for each of the command\-line arguments */

    for (char **ap = argv; *ap != NULL; ap++) {
        printf("\enT: about to mkdir(\e"%s\e")\en", *ap);

        int s = mkdir(*ap, 0700);
        if (s == \-1)
            perror("T: ERROR: mkdir(2)");
        else
            printf("T: SUCCESS: mkdir(2) returned %d\en", s);
    }

    printf("\enT: terminating\en");
    exit(EXIT_SUCCESS);
}

/* Check that the notification ID provided by a SECCOMP_IOCTL_NOTIF_RECV
   operation is still valid. It will no longer be valid if the target
   process has terminated or is no longer blocked in the system call that
   generated the notification (because it was interrupted by a signal).

   This operation can be used when doing such things as accessing
   /proc/PID files in the target process in order to avoid TOCTOU race
   conditions where the PID that is returned by SECCOMP_IOCTL_NOTIF_RECV
   terminates and is reused by another process. */

static bool
cookieIsValid(int notifyFd, uint64_t id)
{
    return ioctl(notifyFd, SECCOMP_IOCTL_NOTIF_ID_VALID, &id) == 0;
}

/* Access the memory of the target process in order to fetch the
   pathname referred to by the system call argument \(aqargNum\(aq in
   \(aqreq\->data.args[]\(aq.  The pathname is returned in \(aqpath\(aq,
   a buffer of \(aqlen\(aq bytes allocated by the caller.

   Returns true if the pathname is successfully fetched, and false
   otherwise. For possible causes of failure, see the comments below. */

static bool
getTargetPathname(struct seccomp_notif *req, int notifyFd,
                  int argNum, char *path, size_t len)
{
    char procMemPath[PATH_MAX];

    snprintf(procMemPath, sizeof(procMemPath), "/proc/%d/mem", req\->pid);

    int procMemFd = open(procMemPath, O_RDONLY | O_CLOEXEC);
    if (procMemFd == \-1)
        return false;

    /* Check that the process whose info we are accessing is still alive
       and blocked in the system call that caused the notification.
       If the SECCOMP_IOCTL_NOTIF_ID_VALID operation (performed in
       cookieIsValid()) succeeded, we know that the /proc/PID/mem file
       descriptor that we opened corresponded to the process for which we
       received a notification. If that process subsequently terminates,
       then read() on that file descriptor will return 0 (EOF). */

    if (!cookieIsValid(notifyFd, req\->id)) {
        close(procMemFd);
        return false;
    }

    /* Read bytes at the location containing the pathname argument */

    ssize_t nread = pread(procMemFd, path, len, req\->data.args[argNum]);

    close(procMemFd);

    if (nread <= 0)
        return false;

    /* Once again check that the notification ID is still valid. The
       case we are particularly concerned about here is that just
       before we fetched the pathname, the target\(aqs blocked system
       call was interrupted by a signal handler, and after the handler
       returned, the target carried on execution (past the interrupted
       system call). In that case, we have no guarantees about what we
       are reading, since the target\(aqs memory may have been arbitrarily
       changed by subsequent operations. */

    if (!cookieIsValid(notifyFd, req\->id)) {
        perror("\etS: notification ID check failed!!!");
        return false;
    }

    /* Even if the target\(aqs system call was not interrupted by a signal,
       we have no guarantees about what was in the memory of the target
       process. (The memory may have been modified by another thread, or
       even by an external attacking process.) We therefore treat the
       buffer returned by pread() as untrusted input. The buffer should
       contain a terminating null byte; if not, then we will trigger an
       error for the target process. */

    if (strnlen(path, nread) < nread)
        return true;

    return false;
}

/* Allocate buffers for the seccomp user\-space notification request and
   response structures. It is the caller\(aqs responsibility to free the
   buffers returned via \(aqreq\(aq and \(aqresp\(aq. */

static void
allocSeccompNotifBuffers(struct seccomp_notif **req,
        struct seccomp_notif_resp **resp,
        struct seccomp_notif_sizes *sizes)
{
    /* Discover the sizes of the structures that are used to receive
       notifications and send notification responses, and allocate
       buffers of those sizes. */

    if (seccomp(SECCOMP_GET_NOTIF_SIZES, 0, sizes) == \-1)
        errExit("seccomp\-SECCOMP_GET_NOTIF_SIZES");

    *req = malloc(sizes\->seccomp_notif);
    if (*req == NULL)
        errExit("malloc\-seccomp_notif");

    /* When allocating the response buffer, we must allow for the fact
       that the user\-space binary may have been built with user\-space
       headers where \(aqstruct seccomp_notif_resp\(aq is bigger than the
       response buffer expected by the (older) kernel. Therefore, we
       allocate a buffer that is the maximum of the two sizes. This
       ensures that if the supervisor places bytes into the response
       structure that are past the response size that the kernel expects,
       then the supervisor is not touching an invalid memory location. */

    size_t resp_size = sizes\->seccomp_notif_resp;
    if (sizeof(struct seccomp_notif_resp) > resp_size)
        resp_size = sizeof(struct seccomp_notif_resp);

    *resp = malloc(resp_size);
    if (resp == NULL)
        errExit("malloc\-seccomp_notif_resp");

}

/* Handle notifications that arrive via the SECCOMP_RET_USER_NOTIF file
   descriptor, \(aqnotifyFd\(aq. */

static void
handleNotifications(int notifyFd)
{
    struct seccomp_notif_sizes sizes;
    struct seccomp_notif *req;
    struct seccomp_notif_resp *resp;
    char path[PATH_MAX];

    allocSeccompNotifBuffers(&req, &resp, &sizes);

    /* Loop handling notifications */

    for (;;) {

        /* Wait for next notification, returning info in \(aq*req\(aq */

        memset(req, 0, sizes.seccomp_notif);
        if (ioctl(notifyFd, SECCOMP_IOCTL_NOTIF_RECV, req) == \-1) {
            if (errno == EINTR)
                continue;
            errExit("\etS: ioctl\-SECCOMP_IOCTL_NOTIF_RECV");
        }

        printf("\etS: got notification (ID %#llx) for PID %d\en",
                req\->id, req\->pid);

        /* The only system call that can generate a notification event
           is mkdir(2). Nevertheless, we check that the notified system
           call is indeed mkdir() as kind of future\-proofing of this
           code in case the seccomp filter is later modified to
           generate notifications for other system calls. */

        if (req\->data.nr != __NR_mkdir) {
            printf("\etS: notification contained unexpected "
                    "system call number; bye!!!\en");
            exit(EXIT_FAILURE);
        }

        bool pathOK = getTargetPathname(req, notifyFd, 0, path,
                                        sizeof(path));

        /* Prepopulate some fields of the response */

        resp\->id = req\->id;     /* Response includes notification ID */
        resp\->flags = 0;
        resp\->val = 0;

        /* If getTargetPathname() failed, trigger an EINVAL error
           response (sending this response may yield an error if the
           failure occurred because the notification ID was no longer
           valid); if the directory is in /tmp, then create it on behalf
           of the supervisor; if the pathname starts with \(aq.\(aq, tell the
           kernel to let the target process execute the mkdir();
           otherwise, give an error for a directory pathname in any other
           location. */

        if (!pathOK) {
            resp->error = -EINVAL;
            printf("\etS: spoofing error for invalid pathname (%s)\en",
                    strerror(-resp->error));
        } else if (strncmp(path, "/tmp/", strlen("/tmp/")) == 0) {
            printf("\etS: executing: mkdir(\e"%s\e", %#llo)\en",
                    path, req\->data.args[1]);

            if (mkdir(path, req\->data.args[1]) == 0) {
                resp\->error = 0;            /* "Success" */
                resp\->val = strlen(path);   /* Used as return value of
                                               mkdir() in target */
                printf("\etS: success! spoofed return = %lld\en",
                        resp\->val);
            } else {

                /* If mkdir() failed in the supervisor, pass the error
                   back to the target */

                resp\->error = \-errno;
                printf("\etS: failure! (errno = %d; %s)\en", errno,
                        strerror(errno));
            }
        } else if (strncmp(path, "./", strlen("./")) == 0) {
            resp\->error = resp\->val = 0;
            resp\->flags = SECCOMP_USER_NOTIF_FLAG_CONTINUE;
            printf("\etS: target can execute system call\en");
        } else {
            resp\->error = \-EOPNOTSUPP;
            printf("\etS: spoofing error response (%s)\en",
                    strerror(\-resp\->error));
        }

        /* Send a response to the notification */

        printf("\etS: sending response "
                "(flags = %#x; val = %lld; error = %d)\en",
                resp\->flags, resp\->val, resp\->error);

        if (ioctl(notifyFd, SECCOMP_IOCTL_NOTIF_SEND, resp) == \-1) {
            if (errno == ENOENT)
                printf("\etS: response failed with ENOENT; "
                        "perhaps target process\(aqs syscall was "
                        "interrupted by a signal?\en");
            else
                perror("ioctl\-SECCOMP_IOCTL_NOTIF_SEND");
        }

        /* If the pathname is just "/bye", then the supervisor breaks out
           of the loop and terminates. This allows us to see what happens
           if the target process makes further calls to mkdir(2). */

        if (strcmp(path, "/bye") == 0)
            break;
    }

    free(req);
    free(resp);
    printf("\etS: terminating **********\en");
    exit(EXIT_FAILURE);
}

/* Implementation of the supervisor process:

   (1) obtains the notification file descriptor from \(aqsockPair[1]\(aq
   (2) handles notifications that arrive on that file descriptor. */

static void
supervisor(int sockPair[2])
{
    int notifyFd = recvfd(sockPair[1]);
    if (notifyFd == \-1)
        errExit("recvfd");

    closeSocketPair(sockPair);  /* We no longer need the socket pair */

    handleNotifications(notifyFd);
}

int
main(int argc, char *argv[])
{
    int sockPair[2];

    setbuf(stdout, NULL);

    if (argc < 2) {
        fprintf(stderr, "At least one pathname argument is required\en");
        exit(EXIT_FAILURE);
    }

    /* Create a UNIX domain socket that is used to pass the seccomp
       notification file descriptor from the target process to the
       supervisor process. */

    if (socketpair(AF_UNIX, SOCK_STREAM, 0, sockPair) == \-1)
        errExit("socketpair");

    /* Create a child process\-\-the "target"\-\-that installs seccomp
       filtering. The target process writes the seccomp notification
       file descriptor onto \(aqsockPair[0]\(aq and then calls mkdir(2) for
       each directory in the command\-line arguments. */

    (void) targetProcess(sockPair, &argv[optind]);

    /* Catch SIGCHLD when the target terminates, so that the
       supervisor can also terminate. */

    struct sigaction sa;
    sa.sa_handler = sigchldHandler;
    sa.sa_flags = 0;
    sigemptyset(&sa.sa_mask);
    if (sigaction(SIGCHLD, &sa, NULL) == \-1)
        errExit("sigaction");

    supervisor(sockPair);

    exit(EXIT_SUCCESS);
}
.EE
.SH SEE ALSO
.BR ioctl (2),
.BR pidfd_open (2),
.BR pidfd_getfd (2),
.BR seccomp (2)
.PP
A further example program can be found in the kernel source file
.IR samples/seccomp/user-trap.c .