============================ KERNEL KEY RETENTION SERVICE ============================ This service allows cryptographic keys, authentication tokens, cross-domain user mappings, and similar to be cached in the kernel for the use of filesystems other kernel services. Keyrings are permitted; these are a special type of key that can hold links to other keys. Processes each have three standard keyring subscriptions that a kernel service can search for relevant keys. The key service can be configured on by enabling: "Security options"/"Enable access key retention support" (CONFIG_KEYS) This document has the following sections: - Key overview - Key service overview - Key access permissions - New procfs files - Userspace system call interface - Kernel services - Defining a key type - Request-key callback service - Key access filesystem ============ KEY OVERVIEW ============ In this context, keys represent units of cryptographic data, authentication tokens, keyrings, etc.. These are represented in the kernel by struct key. Each key has a number of attributes: - A serial number. - A type. - A description (for matching a key in a search). - Access control information. - An expiry time. - A payload. - State. (*) Each key is issued a serial number of type key_serial_t that is unique for the lifetime of that key. All serial numbers are positive non-zero 32-bit integers. Userspace programs can use a key's serial numbers as a way to gain access to it, subject to permission checking. (*) Each key is of a defined "type". Types must be registered inside the kernel by a kernel service (such as a filesystem) before keys of that type can be added or used. Userspace programs cannot define new types directly. Key types are represented in the kernel by struct key_type. This defines a number of operations that can be performed on a key of that type. Should a type be removed from the system, all the keys of that type will be invalidated. (*) Each key has a description. This should be a printable string. The key type provides an operation to perform a match between the description on a key and a criterion string. (*) Each key has an owner user ID, a group ID and a permissions mask. These are used to control what a process may do to a key from userspace, and whether a kernel service will be able to find the key. (*) Each key can be set to expire at a specific time by the key type's instantiation function. Keys can also be immortal. (*) Each key can have a payload. This is a quantity of data that represent the actual "key". In the case of a keyring, this is a list of keys to which the keyring links; in the case of a user-defined key, it's an arbitrary blob of data. Having a payload is not required; and the payload can, in fact, just be a value stored in the struct key itself. When a key is instantiated, the key type's instantiation function is called with a blob of data, and that then creates the key's payload in some way. Similarly, when userspace wants to read back the contents of the key, if permitted, another key type operation will be called to convert the key's attached payload back into a blob of data. (*) Each key can be in one of a number of basic states: (*) Uninstantiated. The key exists, but does not have any data attached. Keys being requested from userspace will be in this state. (*) Instantiated. This is the normal state. The key is fully formed, and has data attached. (*) Negative. This is a relatively short-lived state. The key acts as a note saying that a previous call out to userspace failed, and acts as a throttle on key lookups. A negative key can be updated to a normal state. (*) Expired. Keys can have lifetimes set. If their lifetime is exceeded, they traverse to this state. An expired key can be updated back to a normal state. (*) Revoked. A key is put in this state by userspace action. It can't be found or operated upon (apart from by unlinking it). (*) Dead. The key's type was unregistered, and so the key is now useless. ==================== KEY SERVICE OVERVIEW ==================== The key service provides a number of features besides keys: (*) The key service defines two special key types: (+) "keyring" Keyrings are special keys that contain a list of other keys. Keyring lists can be modified using various system calls. Keyrings should not be given a payload when created. (+) "user" A key of this type has a description and a payload that are arbitrary blobs of data. These can be created, updated and read by userspace, and aren't intended for use by kernel services. (*) Each process subscribes to three keyrings: a thread-specific keyring, a process-specific keyring, and a session-specific keyring. The thread-specific keyring is discarded from the child when any sort of clone, fork, vfork or execve occurs. A new keyring is created only when required. The process-specific keyring is replaced with an empty one in the child on clone, fork, vfork unless CLONE_THREAD is supplied, in which case it is shared. execve also discards the process's process keyring and creates a new one. The session-specific keyring is persistent across clone, fork, vfork and execve, even when the latter executes a set-UID or set-GID binary. A process can, however, replace its current session keyring with a new one by using PR_JOIN_SESSION_KEYRING. It is permitted to request an anonymous new one, or to attempt to create or join one of a specific name. The ownership of the thread keyring changes when the real UID and GID of the thread changes. (*) Each user ID resident in the system holds two special keyrings: a user specific keyring and a default user session keyring. The default session keyring is initialised with a link to the user-specific keyring. When a process changes its real UID, if it used to have no session key, it will be subscribed to the default session key for the new UID. If a process attempts to access its session key when it doesn't have one, it will be subscribed to the default for its current UID. (*) Each user has two quotas against which the keys they own are tracked. One limits the total number of keys and keyrings, the other limits the total amount of description and payload space that can be consumed. The user can view information on this and other statistics through procfs files. Process-specific and thread-specific keyrings are not counted towards a user's quota. If a system call that modifies a key or keyring in some way would put the user over quota, the operation is refused and error EDQUOT is returned. (*) There's a system call interface by which userspace programs can create and manipulate keys and keyrings. (*) There's a kernel interface by which services can register types and search for keys. (*) There's a way for the a search done from the kernel to call back to userspace to request a key that can't be found in a process's keyrings. (*) An optional filesystem is available through which the key database can be viewed and manipulated. ====================== KEY ACCESS PERMISSIONS ====================== Keys have an owner user ID, a group access ID, and a permissions mask. The mask has up to eight bits each for user, group and other access. Only five of each set of eight bits are defined. These permissions granted are: (*) View This permits a key or keyring's attributes to be viewed - including key type and description. (*) Read This permits a key's payload to be viewed or a keyring's list of linked keys. (*) Write This permits a key's payload to be instantiated or updated, or it allows a link to be added to or removed from a keyring. (*) Search This permits keyrings to be searched and keys to be found. Searches can only recurse into nested keyrings that have search permission set. (*) Link This permits a key or keyring to be linked to. To create a link from a keyring to a key, a process must have Write permission on the keyring and Link permission on the key. For changing the ownership, group ID or permissions mask, being the owner of the key or having the sysadmin capability is sufficient. ================ NEW PROCFS FILES ================ Two files have been added to procfs by which an administrator can find out about the status of the key service: (*) /proc/keys This lists all the keys on the system, giving information about their type, description and permissions. The payload of the key is not available this way: SERIAL FLAGS USAGE EXPY PERM UID GID TYPE DESCRIPTION: SUMMARY 00000001 I----- 39 perm 1f0000 0 0 keyring _uid_ses.0: 1/4 00000002 I----- 2 perm 1f0000 0 0 keyring _uid.0: empty 00000007 I----- 1 perm 1f0000 0 0 keyring _pid.1: empty 0000018d I----- 1 perm 1f0000 0 0 keyring _pid.412: empty 000004d2 I--Q-- 1 perm 1f0000 32 -1 keyring _uid.32: 1/4 000004d3 I--Q-- 3 perm 1f0000 32 -1 keyring _uid_ses.32: empty 00000892 I--QU- 1 perm 1f0000 0 0 user metal:copper: 0 00000893 I--Q-N 1 35s 1f0000 0 0 user metal:silver: 0 00000894 I--Q-- 1 10h 1f0000 0 0 user metal:gold: 0 The flags are: I Instantiated R Revoked D Dead Q Contributes to user's quota U Under contruction by callback to userspace N Negative key This file must be enabled at kernel configuration time as it allows anyone to list the keys database. (*) /proc/key-users This file lists the tracking data for each user that has at least one key on the system. Such data includes quota information and statistics: [root@andromeda root]# cat /proc/key-users 0: 46 45/45 1/100 13/10000 29: 2 2/2 2/100 40/10000 32: 2 2/2 2/100 40/10000 38: 2 2/2 2/100 40/10000 The format of each line is <UID>: User ID to which this applies <usage> Structure refcount <inst>/<keys> Total number of keys and number instantiated <keys>/<max> Key count quota <bytes>/<max> Key size quota =============================== USERSPACE SYSTEM CALL INTERFACE =============================== Userspace can manipulate keys directly through three new syscalls: add_key, request_key and keyctl. The latter provides a number of functions for manipulating keys. When referring to a key directly, userspace programs should use the key's serial number (a positive 32-bit integer). However, there are some special values available for referring to special keys and keyrings that relate to the process making the call: CONSTANT VALUE KEY REFERENCED ============================== ====== =========================== KEY_SPEC_THREAD_KEYRING -1 thread-specific keyring KEY_SPEC_PROCESS_KEYRING -2 process-specific keyring KEY_SPEC_SESSION_KEYRING -3 session-specific keyring KEY_SPEC_USER_KEYRING -4 UID-specific keyring KEY_SPEC_USER_SESSION_KEYRING -5 UID-session keyring KEY_SPEC_GROUP_KEYRING -6 GID-specific keyring The main syscalls are: (*) Create a new key of given type, description and payload and add it to the nominated keyring: key_serial_t add_key(const char *type, const char *desc, const void *payload, size_t plen, key_serial_t keyring); If a key of the same type and description as that proposed already exists in the keyring, this will try to update it with the given payload, or it will return error EEXIST if that function is not supported by the key type. The process must also have permission to write to the key to be able to update it. The new key will have all user permissions granted and no group or third party permissions. Otherwise, this will attempt to create a new key of the specified type and description, and to instantiate it with the supplied payload and attach it to the keyring. In this case, an error will be generated if the process does not have permission to write to the keyring. The payload is optional, and the pointer can be NULL if not required by the type. The payload is plen in size, and plen can be zero for an empty payload. A new keyring can be generated by setting type "keyring", the keyring name as the description (or NULL) and setting the payload to NULL. User defined keys can be created by specifying type "user". It is recommended that a user defined key's description by prefixed with a type ID and a colon, such as "krb5tgt:" for a Kerberos 5 ticket granting ticket. Any other type must have been registered with the kernel in advance by a kernel service such as a filesystem. The ID of the new or updated key is returned if successful. (*) Search the process's keyrings for a key, potentially calling out to userspace to create it. key_serial_t request_key(const char *type, const char *description, const char *callout_info, key_serial_t dest_keyring); This function searches all the process's keyrings in the order thread, process, session for a matching key. This works very much like KEYCTL_SEARCH, including the optional attachment of the discovered key to a keyring. If a key cannot be found, and if callout_info is not NULL, then /sbin/request-key will be invoked in an attempt to obtain a key. The callout_info string will be passed as an argument to the program. The keyctl syscall functions are: (*) Map a special key ID to a real key ID for this process: key_serial_t keyctl(KEYCTL_GET_KEYRING_ID, key_serial_t id, int create); The special key specified by "id" is looked up (with the key being created if necessary) and the ID of the key or keyring thus found is returned if it exists. If the key does not yet exist, the key will be created if "create" is non-zero; and the error ENOKEY will be returned if "create" is zero. (*) Replace the session keyring this process subscribes to with a new one: key_serial_t keyctl(KEYCTL_JOIN_SESSION_KEYRING, const char *name); If name is NULL, an anonymous keyring is created attached to the process as its session keyring, displacing the old session keyring. If name is not NULL, if a keyring of that name exists, the process attempts to attach it as the session keyring, returning an error if that is not permitted; otherwise a new keyring of that name is created and attached as the session keyring. To attach to a named keyring, the keyring must have search permission for the process's ownership. The ID of the new session keyring is returned if successful. (*) Update the specified key: long keyctl(KEYCTL_UPDATE, key_serial_t key, const void *payload, size_t plen); This will try to update the specified key with the given payload, or it will return error EOPNOTSUPP if that function is not supported by the key type. The process must also have permission to write to the key to be able to update it. The payload is of length plen, and may be absent or empty as for add_key(). (*) Revoke a key: long keyctl(KEYCTL_REVOKE, key_serial_t key); This makes a key unavailable for further operations. Further attempts to use the key will be met with error EKEYREVOKED, and the key will no longer be findable. (*) Change the ownership of a key: long keyctl(KEYCTL_CHOWN, key_serial_t key, uid_t uid, gid_t gid); This function permits a key's owner and group ID to be changed. Either one of uid or gid can be set to -1 to suppress that change. Only the superuser can change a key's owner to something other than the key's current owner. Similarly, only the superuser can change a key's group ID to something other than the calling process's group ID or one of its group list members. (*) Change the permissions mask on a key: long keyctl(KEYCTL_SETPERM, key_serial_t key, key_perm_t perm); This function permits the owner of a key or the superuser to change the permissions mask on a key. Only bits the available bits are permitted; if any other bits are set, error EINVAL will be returned. (*) Describe a key: long keyctl(KEYCTL_DESCRIBE, key_serial_t key, char *buffer, size_t buflen); This function returns a summary of the key's attributes (but not its payload data) as a string in the buffer provided. Unless there's an error, it always returns the amount of data it could produce, even if that's too big for the buffer, but it won't copy more than requested to userspace. If the buffer pointer is NULL then no copy will take place. A process must have view permission on the key for this function to be successful. If successful, a string is placed in the buffer in the following format: <type>;<uid>;<gid>;<perm>;<description> Where type and description are strings, uid and gid are decimal, and perm is hexadecimal. A NUL character is included at the end of the string if the buffer is sufficiently big. This can be parsed with sscanf(buffer, "%[^;];%d;%d;%o;%s", type, &uid, &gid, &mode, desc); (*) Clear out a keyring: long keyctl(KEYCTL_CLEAR, key_serial_t keyring); This function clears the list of keys attached to a keyring. The calling process must have write permission on the keyring, and it must be a keyring (or else error ENOTDIR will result). (*) Link a key into a keyring: long keyctl(KEYCTL_LINK, key_serial_t keyring, key_serial_t key); This function creates a link from the keyring to the key. The process must have write permission on the keyring and must have link permission on the key. Should the keyring not be a keyring, error ENOTDIR will result; and if the keyring is full, error ENFILE will result. The link procedure checks the nesting of the keyrings, returning ELOOP if it appears to deep or EDEADLK if the link would introduce a cycle. (*) Unlink a key or keyring from another keyring: long keyctl(KEYCTL_UNLINK, key_serial_t keyring, key_serial_t key); This function looks through the keyring for the first link to the specified key, and removes it if found. Subsequent links to that key are ignored. The process must have write permission on the keyring. If the keyring is not a keyring, error ENOTDIR will result; and if the key is not present, error ENOENT will be the result. (*) Search a keyring tree for a key: key_serial_t keyctl(KEYCTL_SEARCH, key_serial_t keyring, const char *type, const char *description, key_serial_t dest_keyring); This searches the keyring tree headed by the specified keyring until a key is found that matches the type and description criteria. Each keyring is checked for keys before recursion into its children occurs. The process must have search permission on the top level keyring, or else error EACCES will result. Only keyrings that the process has search permission on will be recursed into, and only keys and keyrings for which a process has search permission can be matched. If the specified keyring is not a keyring, ENOTDIR will result. If the search succeeds, the function will attempt to link the found key into the destination keyring if one is supplied (non-zero ID). All the constraints applicable to KEYCTL_LINK apply in this case too. Error ENOKEY, EKEYREVOKED or EKEYEXPIRED will be returned if the search fails. On success, the resulting key ID will be returned. (*) Read the payload data from a key: key_serial_t keyctl(KEYCTL_READ, key_serial_t keyring, char *buffer, size_t buflen); This function attempts to read the payload data from the specified key into the buffer. The process must have read permission on the key to succeed. The returned data will be processed for presentation by the key type. For instance, a keyring will return an array of key_serial_t entries representing the IDs of all the keys to which it is subscribed. The user defined key type will return its data as is. If a key type does not implement this function, error EOPNOTSUPP will result. As much of the data as can be fitted into the buffer will be copied to userspace if the buffer pointer is not NULL. On a successful return, the function will always return the amount of data available rather than the amount copied. (*) Instantiate a partially constructed key. key_serial_t keyctl(KEYCTL_INSTANTIATE, key_serial_t key, const void *payload, size_t plen, key_serial_t keyring); If the kernel calls back to userspace to complete the instantiation of a key, userspace should use this call to supply data for the key before the invoked process returns, or else the key will be marked negative automatically. The process must have write access on the key to be able to instantiate it, and the key must be uninstantiated. If a keyring is specified (non-zero), the key will also be linked into that keyring, however all the constraints applying in KEYCTL_LINK apply in this case too. The payload and plen arguments describe the payload data as for add_key(). (*) Negatively instantiate a partially constructed key. key_serial_t keyctl(KEYCTL_NEGATE, key_serial_t key, unsigned timeout, key_serial_t keyring); If the kernel calls back to userspace to complete the instantiation of a key, userspace should use this call mark the key as negative before the invoked process returns if it is unable to fulfil the request. The process must have write access on the key to be able to instantiate it, and the key must be uninstantiated. If a keyring is specified (non-zero), the key will also be linked into that keyring, however all the constraints applying in KEYCTL_LINK apply in this case too. =============== KERNEL SERVICES =============== The kernel services for key managment are fairly simple to deal with. They can be broken down into two areas: keys and key types. Dealing with keys is fairly straightforward. Firstly, the kernel service registers its type, then it searches for a key of that type. It should retain the key as long as it has need of it, and then it should release it. For a filesystem or device file, a search would probably be performed during the open call, and the key released upon close. How to deal with conflicting keys due to two different users opening the same file is left to the filesystem author to solve. When accessing a key's payload data, key->lock should be at least read locked, or else the data may be changed by an update being performed from userspace whilst the driver or filesystem is trying to access it. If no update method is supplied, then the key's payload may be accessed without holding a lock as there is no way to change it, provided it can be guaranteed that the key's type definition won't go away. (*) To search for a key, call: struct key *request_key(const struct key_type *type, const char *description, const char *callout_string); This is used to request a key or keyring with a description that matches the description specified according to the key type's match function. This permits approximate matching to occur. If callout_string is not NULL, then /sbin/request-key will be invoked in an attempt to obtain the key from userspace. In that case, callout_string will be passed as an argument to the program. Should the function fail error ENOKEY, EKEYEXPIRED or EKEYREVOKED will be returned. (*) When it is no longer required, the key should be released using: void key_put(struct key *key); This can be called from interrupt context. If CONFIG_KEYS is not set then the argument will not be parsed. (*) Extra references can be made to a key by calling the following function: struct key *key_get(struct key *key); These need to be disposed of by calling key_put() when they've been finished with. The key pointer passed in will be returned. If the pointer is NULL or CONFIG_KEYS is not set then the key will not be dereferenced and no increment will take place. (*) A key's serial number can be obtained by calling: key_serial_t key_serial(struct key *key); If key is NULL or if CONFIG_KEYS is not set then 0 will be returned (in the latter case without parsing the argument). (*) If a keyring was found in the search, this can be further searched by: struct key *keyring_search(struct key *keyring, const struct key_type *type, const char *description) This searches the keyring tree specified for a matching key. Error ENOKEY is returned upon failure. If successful, the returned key will need to be released. (*) To check the validity of a key, this function can be called: int validate_key(struct key *key); This checks that the key in question hasn't expired or and hasn't been revoked. Should the key be invalid, error EKEYEXPIRED or EKEYREVOKED will be returned. If the key is NULL or if CONFIG_KEYS is not set then 0 will be returned (in the latter case without parsing the argument). (*) To register a key type, the following function should be called: int register_key_type(struct key_type *type); This will return error EEXIST if a type of the same name is already present. (*) To unregister a key type, call: void unregister_key_type(struct key_type *type); =================== DEFINING A KEY TYPE =================== A kernel service may want to define its own key type. For instance, an AFS filesystem might want to define a Kerberos 5 ticket key type. To do this, it author fills in a struct key_type and registers it with the system. The structure has a number of fields, some of which are mandatory: (*) const char *name The name of the key type. This is used to translate a key type name supplied by userspace into a pointer to the structure. (*) size_t def_datalen This is optional - it supplies the default payload data length as contributed to the quota. If the key type's payload is always or almost always the same size, then this is a more efficient way to do things. The data length (and quota) on a particular key can always be changed during instantiation or update by calling: int key_payload_reserve(struct key *key, size_t datalen); With the revised data length. Error EDQUOT will be returned if this is not viable. (*) int (*instantiate)(struct key *key, const void *data, size_t datalen); This method is called to attach a payload to a key during construction. The payload attached need not bear any relation to the data passed to this function. If the amount of data attached to the key differs from the size in keytype->def_datalen, then key_payload_reserve() should be called. This method does not have to lock the key in order to attach a payload. The fact that KEY_FLAG_INSTANTIATED is not set in key->flags prevents anything else from gaining access to the key. This method may sleep if it wishes. (*) int (*duplicate)(struct key *key, const struct key *source); If this type of key can be duplicated, then this method should be provided. It is called to copy the payload attached to the source into the new key. The data length on the new key will have been updated and the quota adjusted already. This method will be called with the source key's semaphore read-locked to prevent its payload from being changed. It is safe to sleep here. (*) int (*update)(struct key *key, const void *data, size_t datalen); If this type of key can be updated, then this method should be provided. It is called to update a key's payload from the blob of data provided. key_payload_reserve() should be called if the data length might change before any changes are actually made. Note that if this succeeds, the type is committed to changing the key because it's already been altered, so all memory allocation must be done first. key_payload_reserve() should be called with the key->lock write locked, and the changes to the key's attached payload should be made before the key is locked. The key will have its semaphore write-locked before this method is called. Any changes to the key should be made with the key's rwlock write-locked also. It is safe to sleep here. (*) int (*match)(const struct key *key, const void *desc); This method is called to match a key against a description. It should return non-zero if the two match, zero if they don't. This method should not need to lock the key in any way. The type and description can be considered invariant, and the payload should not be accessed (the key may not yet be instantiated). It is not safe to sleep in this method; the caller may hold spinlocks. (*) void (*destroy)(struct key *key); This method is optional. It is called to discard the payload data on a key when it is being destroyed. This method does not need to lock the key; it can consider the key as being inaccessible. Note that the key's type may have changed before this function is called. It is not safe to sleep in this method; the caller may hold spinlocks. (*) void (*describe)(const struct key *key, struct seq_file *p); This method is optional. It is called during /proc/keys reading to summarise a key's description and payload in text form. This method will be called with the key's rwlock read-locked. This will prevent the key's payload and state changing; also the description should not change. This also means it is not safe to sleep in this method. (*) long (*read)(const struct key *key, char __user *buffer, size_t buflen); This method is optional. It is called by KEYCTL_READ to translate the key's payload into something a blob of data for userspace to deal with. Ideally, the blob should be in the same format as that passed in to the instantiate and update methods. If successful, the blob size that could be produced should be returned rather than the size copied. This method will be called with the key's semaphore read-locked. This will prevent the key's payload changing. It is not necessary to also read-lock key->lock when accessing the key's payload. It is safe to sleep in this method, such as might happen when the userspace buffer is accessed. ============================ REQUEST-KEY CALLBACK SERVICE ============================ To create a new key, the kernel will attempt to execute the following command line: /sbin/request-key create <key> <uid> <gid> \ <threadring> <processring> <sessionring> <callout_info> <key> is the key being constructed, and the three keyrings are the process keyrings from the process that caused the search to be issued. These are included for two reasons: (1) There may be an authentication token in one of the keyrings that is required to obtain the key, eg: a Kerberos Ticket-Granting Ticket. (2) The new key should probably be cached in one of these rings. This program should set it UID and GID to those specified before attempting to access any more keys. It may then look around for a user specific process to hand the request off to (perhaps a path held in placed in another key by, for example, the KDE desktop manager). The program (or whatever it calls) should finish construction of the key by calling KEYCTL_INSTANTIATE, which also permits it to cache the key in one of the keyrings (probably the session ring) before returning. Alternatively, the key can be marked as negative with KEYCTL_NEGATE; this also permits the key to be cached in one of the keyrings. If it returns with the key remaining in the unconstructed state, the key will be marked as being negative, it will be added to the session keyring, and an error will be returned to the key requestor. Supplementary information may be provided from whoever or whatever invoked this service. This will be passed as the <callout_info> parameter. If no such information was made available, then "-" will be passed as this parameter instead. Similarly, the kernel may attempt to update an expired or a soon to expire key by executing: /sbin/request-key update <key> <uid> <gid> \ <threadring> <processring> <sessionring> In this case, the program isn't required to actually attach the key to a ring; the rings are provided for reference.