man-pages/man2/keyctl.2

.\" Copyright (C) 2016 Michael Kerrisk <mtk.manpages@gmail.com>
.\" and Copyright (C) 2016 Eugene Syromyatnikov <evgsyr@gmail.com>
.\" A very few fragments remain from an earlier version of this page
.\" written by David Howells (dhowells@redhat.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.\" manual page may be incorrect or out-of-date.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 KEYCTL 2 2015-05-07 Linux "Linux Key Management Calls"
.SH NAME
keyctl \- manipulate the kernel's key management facility
.SH SYNOPSIS
.nf
.B #include <sys/types.h>
.B #include <keyutils.h>
.sp
.BI "long keyctl(int " operation ", ...)"
.sp
.B "/* For direct call via syscall(2): */"
.B #include <asm/unistd.h>
.B #include <linux/keyctl.h>
.B #include <unistd.h>
.sp
.BI "long syscall(__NR_keyctl, int " operation ", __kernel_ulong_t " arg2 ,
.BI "             __kernel_ulong_t " arg3 ", __kernel_ulong_t " arg4 ,
.BI "             __kernel_ulong_t " arg5 );
.fi

No glibc wrapper is provided for this system call; see NOTES.
.SH DESCRIPTION
.BR keyctl ()
allows user-space programs to perform key manipulation.

The operation performed by
.BR keyctl ()
is determined by the value of the
.I operation
argument.
Each of these operations is wrapped by
.I libkeyutils
into individual functions (noted below)
to permit the compiler to check types.

The permitted values for
.I operation
are:
.TP
.BR KEYCTL_GET_KEYRING_ID " (since Linux 2.6.11)"
Map a special key ID to a real key ID for this process.

This operation looks up the special key whose ID is provided in
.I arg2
(cast to
.IR key_serial_t )
and (if it is found) the ID of the corresponding real key is returned
as the function result.
The following values may be specified in
.IR arg2 :
.RS
.TP
.B KEY_SPEC_THREAD_KEYRING
This specifies the calling thread's thread-specific keyring.
See
.BR thread-keyring (7).
.TP
.B KEY_SPEC_PROCESS_KEYRING
This specifies the caller's process-specific keyring.
See
.BR process-keyring (7).
.TP
.B KEY_SPEC_SESSION_KEYRING
This specifies the caller's session-specific keyring.
See
.BR session-keyring (7).
.TP
.B KEY_SPEC_USER_KEYRING
This specifies the caller's UID-specific keyring.
See
.BR user-keyring (7).
.TP
.B KEY_SPEC_USER_SESSION_KEYRING
This specifies the caller's UID-session keyring.
See
.BR user-session-keyring (7).
.TP
.BR KEY_SPEC_REQKEY_AUTH_KEY " (since Linux 2.6.16)"
.\"            commit b5f545c880a2a47947ba2118b2509644ab7a2969
This specifies the authorization key created by
.BR request_key (2)
and passed to the process it spawns to generate a key.
This key is available only in a
.BR request-key (8)-style
program that was passed an authorization key by the kernel and
ceases to be available once the requested key has been instantiated; see
.BR request_key (2).
.TP
.BR KEY_SPEC_REQUESTOR_KEYRING " (since Linux 2.6.29)"
.\"            commit 8bbf4976b59fc9fc2861e79cab7beb3f6d647640
This specifies the key ID for the
.BR request_key (2)
destination keyring.
This keyring is available only in a
.BR request-key (8)-style
program that was passed an authorization key by the kernel and
ceases to be available once the requested key has been instantiated; see
.BR request_key (2).
.RE
.IP
If the key specified in
.I arg2
does not exist, then a new key is created if the
.I arg3
argument (cast to
.IR int )
contains a non-zero value; otherwise the operation fails with the error
.BR ENOKEY .
.\" The keyctl_get_keyring_ID.3 page says that a new key
.\" "will be created *if it is appropriate to do so**. What is the
.\" determiner for appropriate?

The caller must have
.I search
permission on a keyring in order for it to be found.

The arguments
.IR arg4
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_get_keyring_ID (3).
.TP
.BR KEYCTL_JOIN_SESSION_KEYRING " (since Linux 2.6.11)"
.\" FIXME What is the use case for joining a new session keyring?
Replace the session keyring this process subscribes to with
a new session keyring.

If
.I arg2
is NULL,
an anonymous keyring with the description "_ses" is created
and the process is subscribed to that keyring as its session keyring,
displacing the previous session keyring.

Otherwise,
.I arg2
(cast to
.IR "char\ *" )
is treated as the description (name) of a keyring,
and the behavior is as follows:
.RS
.IP * 3
If a keyring with a matching description exists,
the process will attempt to subscribe to that keyring if possible;
if that is not possible, an error is returned.
.\" FIXME What error is returned in the above case?
In order to subscribe to the keyring,
the caller must have
.I search
permission on the keyring.
.IP *
If a keyring with a matching description does not exist,
then a new keyring with that description is created,
and the process is subscribed to that keyring as its session keyring,
displacing the previous session keyring.
.RE
.IP
The arguments
.IR arg3 ,
.IR arg4 ,
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_join_session_keyring (3).
.TP
.BR KEYCTL_UPDATE " (since Linux 2.6.11)"
Update a key's data payload.

The
.I arg2
argument (cast to
.IR key_serial_t )
specifies the ID of the key to be updated.
The
.I arg3
argument (cast to
.IR "void\ *" )
points to the new payload and
.I arg4
(cast to
.IR size_t )
contains the new payload size in bytes.

The caller must have
.I write
permission on the key specified and the key type must support updating.

A negatively instantiated key can be positively instantiated
with this operation.

The
.I arg5
argument is ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_update (3).
.TP
.BR KEYCTL_REVOKE " (since Linux 2.6.11)"
Revoke the key with the ID provided in
.I arg2
(cast to
.IR key_serial_t ).
The key will no longer be findable,
and it will be unavailable for further operations.
Further attempts to use the key will fail with the error
.BR EKEYREVOKED .
.\" FIXME Does a revoked key get garbage collected?

The caller must have
.IR write
or
.IR setattr
permission on the key.
.\" FIXME Keys with the KEY_FLAG_KEEP bit set cause an EPERM
.\" error for KEYCTL_REVOKE. Does this need to be documented?
.\" (It's not clear how KEY_FLAG_KEEP gets set.)

The arguments
.IR arg3 ,
.IR arg4 ,
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_revoke (3).
.TP
.BR KEYCTL_CHOWN " (since Linux 2.6.11)"
Change the ownership (user and group ID) of a key.

The
.I arg2
argument (cast to
.IR key_serial_t )
contains the key ID.
The
.I arg3
argument (cast to
.IR uid_t )
contains the new user ID (or \-1 in case the user ID shouldn't be changed).
The
.I arg4
argument (cast to
.IR gid_t )
contains the new group ID (or \-1 in case the group ID shouldn't be changed).

The key must grant the caller
.I setattr
permission.

For the UID to be changed, or for the GID to be changed to a group
the caller is not a member of, the caller must have the
.B CAP_SYS_ADMIN
capability (see
.BR capabilities (7)).

If the UID is to be changed, the new user must have sufficient
quota to accept the key.
The quota deduction will be removed from the old user
to the new user should the UID be changed.

The
.I arg5
argument is ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_chown (3).
.TP
.BR KEYCTL_SETPERM " (since Linux 2.6.11)"
Change the permissions of the key with the ID provided in the
.I arg2
argument (cast to
.IR key_serial_t )
to the permissions provided in the
.I arg3
argument (cast to
.IR key_perm_t ).

The key must grant
.I setattr
permission to the caller.

If the caller doesn't have the
.B CAP_SYS_ADMIN
capability, it can change permissions only for the keys it owns.
(More precisely: the caller's filesystem UID must match the UID of the key.)

The permissions in
.IR arg3
specify masks of available operations
for each of the following user categories:
.RS
.TP
.IR possessor " (since Linux 2.6.14)"
.\" commit 664cceb0093b755739e56572b836a99104ee8a75
This is the permission granted to a process that possesses the key
(has it attached searchably to one of the process's keyrings);
see
.BR keyrings (7).
.TP
.IR user
This is the permission granted to a process
whose filesystem UID matches the UID of the key.
.TP
.IR group
This is the permission granted to a process
whose filesystem GID or any of its supplementary GIDs
matches the GID of the key.
.TP
.IR other
This is the permission granted to other processes
that do not match the
.IR user
and
.IR group
categories.
.RE
.IP
The
.IR user ,
.IR group ,
and
.IR other
categories are exclusive: if a process matches the
.IR user
category, it will not receive permissions granted in the
.IR group
category; if a process matches the
.I user
or
.IR group
category, then it will not receive permissions granted in the
.IR other
category.

The
.I possessor 
category grants permissions that are cumulative with the grants from the
.IR user ,
.IR group ,
or
.IR other
category.

Each permission mask is eight bits in size,
with only six bits currently used.
The available permissions are:
.RS
.TP
.IR view
This permission allows reading attributes of a key.

This permission is required for the
.BR KEYCTL_DESCRIBE
operation.

The permission bits for each category are
.BR KEY_POS_VIEW ,
.BR KEY_USR_VIEW ,
.BR KEY_GRP_VIEW ,
and
.BR KEY_OTH_VIEW .
.TP
.IR read
This permission allows reading a key's payload.

This permission is required for the
.BR KEYCTL_READ
operation.

The permission bits for each category are
.BR KEY_POS_READ ,
.BR KEY_USR_READ ,
.BR KEY_GRP_READ ,
and
.BR KEY_OTH_READ .
.TP
.IR write
This permission allows update or instantiation of a key's payload.
For a keyring, it allows keys to be linked and unlinked from the keyring,

This permission is required for the
.BR KEYCTL_UPDATE ,
.BR KEYCTL_REVOKE ,
.BR KEYCTL_CLEAR ,
.BR KEYCTL_LINK ,
and
.BR KEYCTL_UNLINK
operations.

The permission bits for each category are
.BR KEY_POS_WRITE ,
.BR KEY_USR_WRITE ,
.BR KEY_GRP_WRITE ,
and
.BR KEY_OTH_WRITE .
.TP
.IR search
This permission allows keyrings to be searched and keys to be found.
Searches can recurse only into nested keyrings
that have search permission set.

This permission is required for the
.BR KEYCTL_GET_KEYRING_ID ,
.BR KEYCTL_JOIN_SESSION_KEYRING ,
.BR KEYCTL_SEARCH ,
and
.BR KEYCTL_INVALIDATE
operations.

The permission bits for each category are
.BR KEY_POS_SEARCH ,
.BR KEY_USR_SEARCH ,
.BR KEY_GRP_SEARCH ,
and
.BR KEY_OTH_SEARCH .
.TP
.IR link
This permission allows a key or keyring to be linked to.

This permission is required for the
.BR KEYCTL_LINK
and
.BR KEYCTL_SESSION_TO_PARENT
operations.

The permission bits for each category are
.BR KEY_POS_LINK ,
.BR KEY_USR_LINK ,
.BR KEY_GRP_LINK ,
and
.BR KEY_OTH_LINK .
.TP
.IR setattr " (since Linux 2.6.15)."
This permission allows a key's UID, GID, and permissions mask to be changed.

This permission is required for the
.BR KEYCTL_REVOKE ,
.BR KEYCTL_CHOWN ,
and
.BR KEYCTL_SETPERM
operations.

The permission bits for each category are
.BR KEY_POS_SETATTR ,
.BR KEY_USR_SETATTR ,
.BR KEY_GRP_SETATTR ,
and
.BR KEY_OTH_SETATTR .
.RE
.IP
As a convenience, the following macros are defined as masks for
all of the permission bits in each of the user categories:
.BR KEY_POS_ALL ,
.BR KEY_USR_ALL,
.BR KEY_GRP_ALL ,
and
.BR KEY_OTH_ALL .

The
.IR arg4 " and " arg5
arguments are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_setperm (3).
.TP
.BR KEYCTL_DESCRIBE " (since Linux 2.6.11)"
Obtain a string describing the attributes of a specified key.

The ID of the key to be described is specified in
.I arg2
(cast to
.IR key_serial_t ).
The descriptive string is returned in the buffer pointed to by
.I arg3
(cast to
.IR "char\ *" );
.I arg4
(cast to
.IR size_t )
specifies the size of that buffer in bytes.

The key must grant the caller
.I view
permission.

The returned string is null-terminated and
contains the following information about the key:

.in +4n
.IR type ; uid ; gid ; perm ; description
.in

In the above,
.IR type
and
.IR description
are strings,
.IR uid
and
.IR gid
are decimal strings, and
.I perm
is a hexadecimal permissions mask.
The descriptive string is written with the following format:

    %s;%d;%d;%08x;%s

.BR "Note: the intention is that the descriptive string should"
.BR "be extensible in future kernel versions".
In particular, the
.IR description
field will not contain semicolons;
it should be parsed by working backwards from the end of the string
to find the last semicolon.
This allows future semicolon-delimited fields to be inserted
in the descriptive string in the future.

Writing to the buffer is attempted only when
.IR arg3
is non-NULL and the specified buffer size
is large enough to accept the descriptive string
(including the terminating null byte).
'\" Function commentary says it copies up to buflen bytes, but see the
'\" (buffer && buflen >= ret) condition in keyctl_describe_key() in
'\" security/keyctl.c
In order to determine whether the buffer size was too small,
check to see if the return value of the operation is greater than
.IR arg4 .

The
.I arg5
argument is ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_describe (3).
.TP
.B KEYCTL_CLEAR
Clear the contents of (i.e., unlink all keys from) a keyring.

The ID of the key
(which must be of keyring type)
.\" or the error ENOTDIR results
is provided in
.I arg2
(cast to
.IR key_serial_t ).
.\" According to Documentation/security/keys.txt:
.\"     This function can also be used to clear special kernel keyrings if they
.\"     are appropriately marked if the user has CAP_SYS_ADMIN capability.  The
.\"     DNS resolver cache keyring is an example of this.

The caller must have
.I write
permission on the keyring.

The arguments
.IR arg3 ,
.IR arg4 ,
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_clear (3).
.TP
.BR KEYCTL_LINK " (since Linux 2.6.11)"
Create a link from a keyring to a key.

The key to be linked is specified in
.IR arg2
(cast to
.IR key_serial_t );
the keyring is specified in
.IR arg3
(cast to
.IR key_serial_t ).

If a key with the same type and description is already linked in the keyring,
then that key is displaced from the keyring.

Before creating the link,
the kernel checks the nesting of the keyrings and returns appropriate errors
if the nesting is too deep
.\" KEYRING_SEARCH_MAX_DEPTH 6
or if the link would produce a cycle.
.\" FIXME What is the purpose of limiting the nesting to
.\" KEYRING_SEARCH_MAX_DEPTH (6)?

The caller must have
.I link
permission on the key being added and
.I write
permission on the keyring.

The arguments
.IR arg4
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_link (3).
.TP
.BR KEYCTL_UNLINK " (since Linux 2.6.11)"
Unlink a key from a keyring.

The ID of the key to be unlinked is specified in
.I arg2
(cast to
.IR key_serial_t );
the ID of the keyring from which it is to be unlinked is specified in
.I arg3
(cast to
.IR key_serial_t ).

If the key is not currently linked into the keyring, an error results.

The caller must have
.I write
permission on the keyring from which the key is being removed.

If the last link to a key is removed,
then that key will be scheduled for destruction.

The arguments
.IR arg4
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_unlink (3).
.TP
.BR KEYCTL_SEARCH " (since Linux 2.6.11)"
Search for a key in a keyring tree,
returning its ID and optionally linking it to a specified keyring.

The tree to be searched is specified by passing
the ID of the head keyring in
.IR arg2
(cast to
.IR key_serial_t ).
The search is performed breadth-first and recursively.

The
.I arg3
and
.I arg4
arguments specify the key to be searched for:
.I arg3
(cast as
.IR "char\ *" )
contains the key type
(a null-terminated character string up to 32 bytes in size,
including the terminating null byte), and
.I arg4
(cast as
.IR "char\ *" )
contains the description of the key
(a null-terminated character string up to 4096 bytes in size,
including the terminating null byte).

The source keyring must grant
.I search
permission to the caller.
When performing the recursive search, only keyrings that grant the caller
.I search
permission will be searched.
Only keys with for which the caller has
.I search
permission can be found.

If the key is found, its ID is returned as the function result.

If the key is found and
.I arg5
(cast to
.IR key_serial_t )
is nonzero, then, subject to the same constraints and rules as
.BR KEYCTL_LINK ,
the key is linked into the keyring whose ID is specified in
.IR arg5 .
If the destination keyring specified in
.I arg5
already contains a link to a key that has the same type and description,
then that link will be displaced by a link to
the key found by this operation.

Instead of valid existing keyring IDs, the source
.RI ( arg2 )
and destination
.RI ( arg5 )
keyrings can be one of the special keyring IDs listed under
.BR KEYCTL_GET_KEYRING_ID .
.IP
This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_search (3).
.TP
.BR KEYCTL_READ " (since Linux 2.6.11)"
Read the payload data of a key.

The ID of the key whose payload is to be read is specified in
.I arg2
(cast to
.IR key_serial_t ).
This can be the ID of an existing key,
or any of the special key IDs listed for
.BR KEYCTL_GET_KEYRING_ID .
.\" including KEY_SPEC_REQKEY_AUTH_KEY

The payload is placed in the buffer pointed by
.I arg3
(cast to
.IR "char\ *" );
the size of that buffer must be specified in
.I arg4
(cast to
.IR size_t ).

The returned data will be processed for presentation
according to the key type.
For example, a keyring will return an array of
.I key_serial_t
entries representing the IDs of all the keys that are linked to it.
The
.IR "user"
key type will return its data as is.
If a key type does not implement this function,
the operation fails with the error
.BR EOPNOTSUPP .

If
.I arg3
is not NULL,
as much of the payload data as will fit is copied into the buffer.
On a successful return,
the return value is always the total size of the payload data.
To determine whether the buffer was of sufficient size,
check to see that the return value is less than or equal to
the value supplied in
.IR arg4 .

The key must either grant the caller
.I read
permission, or grant the caller
.I search
permission when searched for from the process keyrings.
.\" FIXME Above, what does "searched for from the process keyrings" mean?

The
.I arg5
argument is ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_read (3).
.TP
.BR KEYCTL_INSTANTIATE " (since Linux 2.6.11)"
(Positively) instantiate an uninstantiated key with a specified payload.
.\" FIXME Is the only use for KEYCTL_INSTANTIATE inside a
.\" request-keys(8)-style program?

The ID of the key to be instantiated is provided in
.I arg2
(cast to
.IR key_serial_t ).

The key payload is specified in the buffer pointed to by
.I arg3
(cast to
.IR "void\ *");
the size of that buffer is specified in
.I arg4
(cast to
.IR size_t ).

The payload may be a NULL pointer and the buffer size may be 0
if this is supported by the key type (e.g., it is a keyring).

The operation may be fail if the payload data is in the wrong format
or is otherwise invalid.

If
.I arg5
(cast to
.IR key_serial_t )
is nonzero, then, subject to the same constraints and rules as
.BR KEYCTL_LINK ,
the instantiated key is linked into the keyring whose ID specified in
.IR arg5 .

The caller must have the appropriate authorization key;
see
.BR request_key (2).

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_instantiate (3).
.TP
.BR KEYCTL_NEGATE " (since Linux 2.6.11)"
Negatively instantiate an uninstantiated key.
.\" FIXME Is the only use for KEYCTL_NEGATE inside a
.\" request-keys(8)-style program?

This operation is equivalent to the call:

    keyctl(KEYCTL_REJECT, arg2, arg3, ENOKEY, arg4);

The
.I arg5
argument is ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_negate (3).
.TP
.BR KEYCTL_SET_REQKEY_KEYRING " (since Linux 2.6.13)"
Set the default keyring to which implicitly requested keys
.\" FIXME What are implicitly requested keys? In what circumstances is
.\" KEYCTL_SET_REQKEY_KEYRING useful?
.\"
.\" Are implicit requests just the ones that use the kernel-internal
.\" request_key() function (which is not the same as the request_key(2)
.\" system call)?
.\"
.\" Does this operation have any effect for the request_key(2) system call?
will be linked for this thread, and return the previous setting.
Implicit key requests can occur when, for example, opening files
on an AFS or NFS filesystem.

The
.I arg2
argument (cast to
.IR int )
should contain one of the following values,
to specify the new default keyring:
.RS
.TP
.BR KEY_REQKEY_DEFL_NO_CHANGE
Don't change the default keyring.
This option is useful, since the call returns the current default keyring
(without changing it).
.TP
.BR KEY_REQKEY_DEFL_DEFAULT
This selects the default behaviour,
which is to use the thread-specific keyring if there is one,
otherwise the process-specific keyring if there is one,
otherwise the session keyring if there is one,
otherwise the UID-specific session keyring.
.TP
.BR KEY_REQKEY_DEFL_THREAD_KEYRING
Use the thread-specific keyring
.RB ( thread-keyring (7))
as the new default keyring.
.TP
.BR KEY_REQKEY_DEFL_PROCESS_KEYRING
Use the process-specific keyring
.RB ( process-keyring (7))
as the new default keyring.
.TP
.TP
.BR KEY_REQKEY_DEFL_SESSION_KEYRING
Use the session-specific keyring
.RB ( session-keyring (7))
as the new default keyring.
.TP
.BR KEY_REQKEY_DEFL_USER_KEYRING
Use the UID-specific keyring
.RB ( user-keyring (7))
as the new default keyring.
.TP
.BR KEY_REQKEY_DEFL_USER_SESSION_KEYRING
Use the UID-specific session keyring
.RB ( user_session-keyring (7))
as the new default keyring.
.TP
.BR KEY_REQKEY_DEFL_REQUESTOR_KEYRING " (since Linux 2.6.29)"
'\" 8bbf4976b59fc9fc2861e79cab7beb3f6d647640
Use the requestor keyring.
.\" FIXME The preceding explanation needs to be expanded.
.\" Is the following correct:
.\"
.\" The requestor keyring is the dest_keyring that was supplied
.\" to a call to request_key(2)?
.RE
.IP
All other values are invalid.
.\" (including the still-unsupported KEY_REQKEY_DEFL_GROUP_KEYRING)

The arguments
.IR arg3 ,
.IR arg4 ,
and
.IR arg5
are ignored.

The setting controlled by this operation is inherited by the child of
.BR fork (2)
and preserved across
.BR  execve (2).

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_set_reqkey_keyring (3).
.TP
.BR KEYCTL_SET_TIMEOUT " (since Linux 2.6.16)"
Set a timeout on a key.
.\" FIXME Other than looking in /proc/keys, is there any way of
.\" discovering the timeout on a key?

The ID of the key is specified in
.I arg2
(cast to
.IR key_serial_t ).
The timeout value, in seconds from the current time,
is specified in
.I arg3
(cast to
.IR "unsigned int" ).
The timeout is measured against the realtime clock.

Specifying the timeout value as 0 clears any existing timeout on the key.

The
.I /proc/keys
file displays the remaining time until each key will expire.
(This is the only method of discovering the timeout on a key.)

The caller must either have the
.I setattr
permission on the key
or hold an instantiation authorization token for the key (see
.BR request_key (2)).

The key and any links to the key will be
automatically garbage collected after the timeout expires.
Subsequent attempts to access the key will then fail with the error
.BR EKEYEXPIRED .

This operation cannot be used to set timeouts on negative, revoked,
or expired keys.

The arguments
.IR arg4
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_set_timeout (3).
.TP
.BR KEYCTL_ASSUME_AUTHORITY " (since Linux 2.6.16)"
Assume (or divest) the authority for the calling thread
to instantiate a key.
.\" FIXME Is the only use for KEYCTL_ASSUME_AUTHORITY inside a
.\" request-keys(8)-style program?

The
.I arg2
argument (cast to
.IR key_serial_t )
specifies either a nonzero key ID to assume authority,
or the value 0 to divest authority.

If
.I arg2
is nonzero, then it specifies the ID of an uninstantiated key for which
authority is to be assumed.
That key can then be instantiated using one of
.BR KEYCTL_INSTANTIATE ,
.BR KEYCTL_INSTANTIATE_IOV ,
.BR KEYCTL_REJECT ,
or
.BR KEYCTL_NEGATE .
Once the key has been instantiated,
the thread is automatically divested of authority to instantiate the key.

Authority over a key can be assumed only if the calling thread has present
in its keyrings the authorization key that is
associated with the specified key.
The caller must have
.I search
permission on the authorization key.

If the specified key has a matching authorization key,
then the ID of that key is returned.
The authorization key can be read
.RB ( KEYCTL_READ )
to obtain the callout information passed to
.BR request_key (2).

If the ID given in
.I arg2
is 0, then the currently assumed authority is cleared (divested),
and the value 0 is returned.

The
.BR KEYCTL_ASSUME_AUTHORITY
mechanism allows a program such as
.BR request-key (8)
to assume the necessary authority to instantiate a new uninstantiated key
that was created as a consequence of a call to
.BR request_key (2).
For further information, see
.BR request_key (2)
and the kernel source file
.IR Documentation/security/keys-request-key.txt .

The arguments
.IR arg3 ,
.IR arg4 ,
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_assume_authority (3).
.TP
.BR KEYCTL_GET_SECURITY " (since Linux 2.6.26)"
.\" commit 70a5bb72b55e82fbfbf1e22cae6975fac58a1e2d
Get the LSM (Linux Security Module) security label of the specified key.

The ID of the key whose security label is to be fetched is specified in
.I arg2
(cast to
.IR key_serial_t ).
The security label (terminated by a null byte)
will be placed in the buffer pointed to by
.I arg3
argument (cast to
.IR "char\ *" );
the size of the buffer must be provided in
.I arg4
(cast to
.IR size_t ).

If
.I arg3
is specified as NULL or the buffer size specified in
.IR arg4
is too small, the full size of the security label string
(including the terminating null byte)
is returned as the function result,
and nothing is copied to the buffer.

The caller must have
.I view
permission on the specified key.

The returned security label string will be rendered in a form appropriate
to the LSM in force.
For example, with SELinux, it may look like:

    unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023

If no LSM is currently in force,
then an empty string is placed in the buffer.

The
.I arg5
argument is ignored.

This operation is exposed by
.I libkeyutils
via the functions
.BR keyctl_get_security (3)
and
.BR keyctl_get_security_alloc (3).
.TP
.BR KEYCTL_SESSION_TO_PARENT " (since Linux 2.6.32)"
.\" commit ee18d64c1f632043a02e6f5ba5e045bb26a5465f
Replace the session keyring to which the
.I parent
of the calling process
subscribes with the session keyring of the calling process.
.\" What is the use case for KEYCTL_SESSION_TO_PARENT?
.\" David Howells: the Process Authentication Groups people requested this,
.\" but then didn't use it; maybe there are no users.

The keyring will be replaced in the parent process at the point
where the parent next transitions from kernel space to user space.

The keyring must exist and must grant the caller
.I link
permission.
The parent process must be single-threaded and have
the same effective ownership as this process
and must not be be set-user-ID or set-group-ID.
The UID of the parent process's existing session keyring (f it has one),
as well as the UID of the caller's session keyring
much match the caller's effective UID.

The fact that it is the parent process that is affected by this operation
allows a program such as the shell to start a child process that
uses this operation to change the shell's session keyring.
(This is what the
.BR keyctl (1)
.B new_session
command does.)

The arguments
.IR arg2 ,
.IR arg3 ,
.IR arg4 ,
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_session_to_parent (3).
.TP
.BR KEYCTL_REJECT " (since Linux 2.6.39)"
.\" commit fdd1b94581782a2ddf9124414e5b7a5f48ce2f9c
.\" FIXME We need some text here on why it is useful to negatively
.\" instantiate a key
Mark a key as negatively instantiated and set an expiration timer
on the key.
This operation provides a superset of the functionality of the earlier
.BR KEYCTL_NEGATE
operation.
.\" FIXME Is the only use for KEYCTL_REJECT inside a
.\" request-keys(8)-style program?

The ID of the key that is to be negatively instantiated is specified in
.I arg2
(cast to
.IR key_serial_t ).
The
.I arg3
(cast to
.IR "unsigned int" )
argument specifies the lifetime of the key, in seconds.
The
.I arg4
argument (cast to
.IR "unsigned int" )
specifies the error to be returned when a search hits this key;
typically, this is one of
.BR EKEYREJECTED ,
.BR EKEYREVOKED ,
or
.BR EKEYEXPIRED .

If
.I arg5
(cast to
.IR key_serial_t )
is nonzero, then, subject to the same constraints and rules as
.BR KEYCTL_LINK ,
the negatively instantiated key is linked into the keyring
whose ID specified in
.IR arg5 .

The caller must have the appropriate instantiation permit set
(authorization key, see
.B KEYCTL_ASSUME_AUTHORITY
command and
.BR request_key (2)).

Negative keys are used to rate limit repeated
.BR request_key (2)
calls by causing them to return the error specified until the negative key
expires.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_reject (3).
.TP
.BR KEYCTL_INSTANTIATE_IOV " (since Linux 2.6.39)"
.\" commit ee009e4a0d4555ed522a631bae9896399674f063
Instantiate an uninstantiated key with a payload specified
via a vector of buffers.

This operation is the same as
.BR KEYCTL_INSTANTIATE ,
but the payload data is specified as an array of
.IR iovec
structures:

.in +4n
.nf
struct iovec {
    void  *iov_base;    /* Starting address of buffer */
    size_t iov_len;     /* Size of buffer (in bytes) */
};
.fi
.in

The pointer to the payload vector is specified in
.IR arg3
(cast as
.IR "const struct iovec\ *" ).
The number of items in the vector is specified in
.IR arg4
(cast as
.IR "unsigned int" ).

The
.I arg2
(key ID)
and
.I arg5
(keyring ID)
are interpreted as for
.BR KEYCTL_INSTANTIATE .

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_instantiate_iov (3).
.TP
.BR KEYCTL_INVALIDATE " (since Linux 3.5)"
.\" commit fd75815f727f157a05f4c96b5294a4617c0557da
Mark a key as invalid.
.\" FIXME What is the difference between revoking a key and invalidating a key?

The ID of the key to be invalidated is specified in
.I arg2
(cast to
.IR key_serial_t ).

To invalidate a key,
the caller must have
.I search
permission on the key.
.\" CAP_SYS_ADMIN is permitted to invalidate certain special keys

This operation immediately marks the key as invalid
and schedules garbage collection.
The garbage collector removes the invalidated key from all keyrings and
deletes the key when its reference count reaches zero.
After this operation,
the key will be ignored by all searches,
even if it is not yet deleted.

Keys that are marked invalid become invisible to normal key operations
immediately, though they are still visible in
.I /proc/keys
(marked with an 'i' flag)
until they are actually removed.

The arguments
.IR arg3 ,
.IR arg4 ,
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_invalidate (3).
.TP
.BR KEYCTL_GET_PERSISTENT " (since Linux 3.13)"
.\" commit f36f8c75ae2e7d4da34f4c908cebdb4aa42c977e
Get the persistent keyring
.RB ( persistent-keyring (7))
for a specified user and link it to a specified keyring.
.\" FIXME What is the difference between the user keyring and
.\" the persistent keyring?
.\"
.\" FIXME What is the lifetime of the persistent keyring?
.\"
.\" FIXME The session-keyring(7) page has the following text.
.\" What are the circumstances that dictate the choice?
.\"
.\"     If a process doesn't have a session keyring
.\"     when it is accessed,  then,  under  certain
.\"     circumstances,  the  user  session  keyring
.\"     will be attached as the session keyring and
.\"     under  others a new session keyring will be
.\"     created.

The user ID is specified in
.I arg2
(cast to
.IR uid_t ).
If the value \-1 is specified, the caller's real user ID is used.
The ID of the destination keyring is specified in
.I arg3
(cast to
.IR key_serial_t ).

The caller must have the
.BR CAP_SETUID
capability in its user namespace in order to fetch the persistent keyring
for a user ID that does not match either the real or effective user ID
of the caller.

If the call is successful,
a link to the persistent keyring is added to the keyring
whose ID was specified in
.IR arg3 .

The caller must have
.I write
permission on the keyring.

The persistent keyring will be created by the kernel
if it does not yet exist.

Each time the
.B KEYCTL_GET_PERSISTENT
operation is performed, the persistent keyring will
have its expiration timeout reset to the value in:

   /proc/sys/kernel/keys/persistent_keyring_expiry

Should the timeout be reached,
the persistent keyring will be removed and
everything it pins can then be garbage collected.

Persistent keyrings were added to Linux in kernel version 3.13.

The arguments
.IR arg4
and
.IR arg5
are ignored.

This operation is exposed by
.I libkeyutils
via the function
.BR keyctl_get_persistent (3).
.TP
.BR KEYCTL_DH_COMPUTE " (since Linux 4.7)"
.\" commit ddbb41148724367394d0880c516bfaeed127b52e
Compute a Diffie-Hellman shared secret or public key.
.\" FIXME What is the use case for KEYCTL_DH_COMPUTE?

The
.I arg2
argument is a pointer to a set of parameters containing
serial numbers for three
.IR """user"""
keys used in the Diffie-Hellman calculation,
packaged in a structure of the following form:

.nf
.in +4n
struct keyctl_dh_params {
    int32_t private; /* The local private key */
    int32_t prime; /* The prime, known to both parties */
    int32_t base;  /* The base integer: either a shared
                      generator or the remote public key */
};
.in
.fi

Each of the three keys specified in this structure must grant the caller
.I read
permission.
The payloads of these keys are used to calculate the Diffie-Hellman
result as:

    base ^ private mod prime

If the base is the shared generator, the result is the local public key.
If the base is the remote public key, the result is the shared secret.

The
.I arg3
argument (cast to
.IR "char\ *" )
points to a buffer where the result of the calculation is placed.
The size of that buffer is specified in
.I arg4
(cast to
.IR size_t ).

The buffer must be large enough to accommodate the output data,
otherwise an error is returned.
If
.I arg4
is specified zero,
in which case the buffer is not used and
the operation returns the minimum required buffer size
(i.e., the length of the prime).

The
.I arg5
argument is reserved and must be 0.
.SH RETURN VALUE
For a successful call, the return value depends on the operation:
.TP
.B KEYCTL_GET_KEYRING_ID
The ID of the requested keyring.
.TP
.B KEYCTL_JOIN_SESSION_KEYRING
The ID of the joined session keyring.
.TP
.B KEYCTL_DESCRIBE
The size of the description (including the terminating null byte),
irrespective of the provided buffer size.
.TP
.B KEYCTL_SEARCH
The ID of the key that was found.
.TP
.B KEYCTL_READ
The amount of data that is available in the key,
irrespective of the provided buffer size.
.TP
.B KEYCTL_SET_REQKEY_KEYRING
The ID of the previous default keyring
to which implicitly requested keys were linked
(one of
.BR KEY_REQKEY_DEFL_USER_* ).
.TP
.B KEYCTL_ASSUME_AUTHORITY
Either 0, if the ID given was 0,
or the ID of the authorization key matching the specified key,
if a non-zero key ID was provided.
.TP
.B KEYCTL_GET_SECURITY
The size of the LSM security label string
(including the terminating null byte),
irrespective of the provided buffer size.
.TP
.B KEYCTL_GET_PERSISTENT
The ID of the persistent keyring.
.TP
.B KEYCTL_DH_COMPUTE
The number of bytes copied to the buffer, or, if
.I arg4
is 0, the required buffer size.
.TP
All other commands
Zero.
.PP
On error, \-1 is returned, and
.I errno
is set appropriately to indicate the error.
.SH ERRORS
.TP
.B EACCES
The requested operation wasn't permitted.
.TP
.B EDEADLK
.I operation
is
.BR KEYCTL_LINK
and the requested link would result in a cycle.
.TP
.B EDQUOT
The key quota for the caller's user would be exceeded by creating a key or
linking it to the keyring.
.TP
.B EINVAL
.I operation
was
.B KEYCTL_SETPERM
and an invalid permission bit was specified in
.IR arg3 .
.TP
.B EINVAL
.I operation
was
.BR KEYCTL_SEARCH
and the size of the description in
.IR arg4
(including the terminating null byte) exceeded 4096 bytes.
size of the string (including the terminating null byte) specified in
.I arg3
(the key type)
or
.I arg4
(the key description)
exceeded the limit (32 bytes and 4096 bytes respectively).
.B EINVAL
.I operation
was
.B KEYCTL_DH_COMPUTE
and the buffer size provided is not enough to hold the result.
Provide 0 as a buffer size in order to obtain the minimum buffer size.
.TP
.B EKEYEXPIRED
An expired key was found or specified.
.TP
.B EKEYREJECTED
A rejected key was found or specified.
.TP
.B EKEYREVOKED
A revoked key was found or specified.
.TP
.B ELOOP
.I operation
is
.BR KEYCTL_LINK
and the requested link would cause the maximum nesting depth
for keyrings to be exceeded.
.TP
.BR ENFILE " (Linux kernels before 3.13)"
.\" FIXME Does this error really occur? I could not find where
.\" in the kernel source it is generated, but have not tested
.\" this case from a user-space program
.IR operation
is
.BR KEYCTL_LINK
and the keyring is full.
(Before Linux 3.13,
.\" commit b2a4df200d570b2c33a57e1ebfa5896e4bc81b69
the available space for storing keyring links was limited to
a single page of memory; since Linux 3.13, there is no fixed limit.)
.TP
.B ENOENT
.I operation
is
.B KEYCTL_UNLINK
and the key to be unlinked isn't linked to the keyring.
.TP
.B ENOKEY
No matching key was found or an invalid key was specified.
.TP
.B ENOKEY
The value
.B KEYCTL_GET_KEYRING_ID
was specified in
.IR operation ,
the key specified in
.I arg2
did not exist, and
.I arg3
was zero (meaning don't create the key if it didn't exist).
.TP
.B ENOTDIR
A key of keyring type was expected but the ID of a key with
a different type was provided.
.TP
.B EOPNOTSUPP
.I operation
is
.B KEYCTL_READ
and the key type does not support reading
(e.g., the type is
.IR """login""" ).
.TP
.B EOPNOTSUPP
.I operation
is
.B KEYCTL_UPDATE
and the key type does not support updating.
.TP
.B EPERM
.I operation
was
.BR KEYCTL_GET_PERSISTENT ,
.I arg2
specified a UID other than the calling thread's real or effective UID,
and the caller did not have the
.B CAP_SETUID
capability.
.TP
.B EPERM
.I operation
was
.BR KEYCTL_SESSION_TO_PARENT
and either:
all of the UIDs (GIDs) of the parent process do not match
the effective UID (GID) of the calling process;
the UID of the parent's existing session keyring or
the UID of the caller's session keyring did not match
the effective UID of the caller;
the parent process is not single-thread;
or the parent process is
.BR init (1)
or a kernel thread.
.SH VERSIONS
This system call first appeared in Linux 2.6.11.
.SH CONFORMING TO
This system call is a nonstandard Linux extension.
.SH NOTES
No wrapper for this system call is provided in glibc.
A wrapper is provided in the
.IR libkeyutils
package.
When employing the wrapper in that library, link with
.IR \-lkeyutils .
However, rather than using this system call directly,
you probably want to use the various library functions
mentioned in the descriptions of individual operations above.
.SH SEE ALSO
.ad l
.nh
.BR keyctl (1),
.BR add_key (2),
.BR request_key (2),
.\"     .BR find_key_by_type_and_name (3)
.\"     There is a man page, but this function seems not to exist
.BR keyctl (3),
.BR keyctl_assume_authority (3),
.BR keyctl_chown (3),
.BR keyctl_clear (3),
.BR keyctl_describe (3),
.BR keyctl_describe_alloc (3),
.BR keyctl_get_keyring_ID (3),
.BR keyctl_get_persistent (3),
.BR keyctl_get_security (3),
.BR keyctl_get_security_alloc (3),
.BR keyctl_instantiate (3),
.BR keyctl_instantiate_iov (3),
.BR keyctl_invalidate (3),
.BR keyctl_join_session_keyring (3),
.BR keyctl_link (3),
.BR keyctl_negate (3),
.BR keyctl_read (3),
.BR keyctl_read_alloc (3),
.BR keyctl_reject (3),
.BR keyctl_revoke (3),
.BR keyctl_search (3),
.BR keyctl_session_to_parent (3),
.BR keyctl_set_reqkey_keyring (3),
.BR keyctl_set_timeout (3),
.BR keyctl_setperm (3),
.BR keyctl_unlink (3),
.BR keyctl_update (3),
.BR recursive_key_scan (3),
.BR recursive_session_key_scan (3),
.BR keyrings (7),
.BR keyutils (7),
.BR persistent\-keyring (7),
.BR process\-keyring (7),
.BR session\-keyring (7),
.BR thread\-keyring (7),
.BR user\-keyring (7),
.BR user\-session\-keyring (7),
.BR request\-key (8)

The kernel source files
.IR Documentation/security/keys.txt 
and
.IR Documentation/security/keys\-request\-key.txt .