1============================ 2Kernel Key Retention Service 3============================ 4 5This service allows cryptographic keys, authentication tokens, cross-domain 6user mappings, and similar to be cached in the kernel for the use of 7filesystems and other kernel services. 8 9Keyrings are permitted; these are a special type of key that can hold links to 10other keys. Processes each have three standard keyring subscriptions that a 11kernel service can search for relevant keys. 12 13The key service can be configured on by enabling: 14 15 "Security options"/"Enable access key retention support" (CONFIG_KEYS) 16 17This document has the following sections: 18 19.. contents:: :local: 20 21 22Key Overview 23============ 24 25In this context, keys represent units of cryptographic data, authentication 26tokens, keyrings, etc.. These are represented in the kernel by struct key. 27 28Each key has a number of attributes: 29 30 - A serial number. 31 - A type. 32 - A description (for matching a key in a search). 33 - Access control information. 34 - An expiry time. 35 - A payload. 36 - State. 37 38 39 * Each key is issued a serial number of type key_serial_t that is unique for 40 the lifetime of that key. All serial numbers are positive non-zero 32-bit 41 integers. 42 43 Userspace programs can use a key's serial numbers as a way to gain access 44 to it, subject to permission checking. 45 46 * Each key is of a defined "type". Types must be registered inside the 47 kernel by a kernel service (such as a filesystem) before keys of that type 48 can be added or used. Userspace programs cannot define new types directly. 49 50 Key types are represented in the kernel by struct key_type. This defines a 51 number of operations that can be performed on a key of that type. 52 53 Should a type be removed from the system, all the keys of that type will 54 be invalidated. 55 56 * Each key has a description. This should be a printable string. The key 57 type provides an operation to perform a match between the description on a 58 key and a criterion string. 59 60 * Each key has an owner user ID, a group ID and an ACL. These are used to 61 control what a process may do to a key from userspace, and whether a 62 kernel service will be able to find the key. 63 64 * Each key can be set to expire at a specific time by the key type's 65 instantiation function. Keys can also be immortal. 66 67 * Each key can have a payload. This is a quantity of data that represent the 68 actual "key". In the case of a keyring, this is a list of keys to which 69 the keyring links; in the case of a user-defined key, it's an arbitrary 70 blob of data. 71 72 Having a payload is not required; and the payload can, in fact, just be a 73 value stored in the struct key itself. 74 75 When a key is instantiated, the key type's instantiation function is 76 called with a blob of data, and that then creates the key's payload in 77 some way. 78 79 Similarly, when userspace wants to read back the contents of the key, if 80 permitted, another key type operation will be called to convert the key's 81 attached payload back into a blob of data. 82 83 * Each key can be in one of a number of basic states: 84 85 * Uninstantiated. The key exists, but does not have any data attached. 86 Keys being requested from userspace will be in this state. 87 88 * Instantiated. This is the normal state. The key is fully formed, and 89 has data attached. 90 91 * Negative. This is a relatively short-lived state. The key acts as a 92 note saying that a previous call out to userspace failed, and acts as 93 a throttle on key lookups. A negative key can be updated to a normal 94 state. 95 96 * Expired. Keys can have lifetimes set. If their lifetime is exceeded, 97 they traverse to this state. An expired key can be updated back to a 98 normal state. 99 100 * Revoked. A key is put in this state by userspace action. It can't be 101 found or operated upon (apart from by unlinking it). 102 103 * Dead. The key's type was unregistered, and so the key is now useless. 104 105Keys in the last three states are subject to garbage collection. See the 106section on "Garbage collection". 107 108 109Key Service Overview 110==================== 111 112The key service provides a number of features besides keys: 113 114 * The key service defines three special key types: 115 116 (+) "keyring" 117 118 Keyrings are special keys that contain a list of other keys. Keyring 119 lists can be modified using various system calls. Keyrings should not 120 be given a payload when created. 121 122 (+) "user" 123 124 A key of this type has a description and a payload that are arbitrary 125 blobs of data. These can be created, updated and read by userspace, 126 and aren't intended for use by kernel services. 127 128 (+) "logon" 129 130 Like a "user" key, a "logon" key has a payload that is an arbitrary 131 blob of data. It is intended as a place to store secrets which are 132 accessible to the kernel but not to userspace programs. 133 134 The description can be arbitrary, but must be prefixed with a non-zero 135 length string that describes the key "subclass". The subclass is 136 separated from the rest of the description by a ':'. "logon" keys can 137 be created and updated from userspace, but the payload is only 138 readable from kernel space. 139 140 * Each process subscribes to three keyrings: a thread-specific keyring, a 141 process-specific keyring, and a session-specific keyring. 142 143 The thread-specific keyring is discarded from the child when any sort of 144 clone, fork, vfork or execve occurs. A new keyring is created only when 145 required. 146 147 The process-specific keyring is replaced with an empty one in the child on 148 clone, fork, vfork unless CLONE_THREAD is supplied, in which case it is 149 shared. execve also discards the process's process keyring and creates a 150 new one. 151 152 The session-specific keyring is persistent across clone, fork, vfork and 153 execve, even when the latter executes a set-UID or set-GID binary. A 154 process can, however, replace its current session keyring with a new one 155 by using PR_JOIN_SESSION_KEYRING. It is permitted to request an anonymous 156 new one, or to attempt to create or join one of a specific name. 157 158 The ownership of the thread keyring changes when the real UID and GID of 159 the thread changes. 160 161 * Each user ID resident in the system holds two special keyrings: a user 162 specific keyring and a default user session keyring. The default session 163 keyring is initialised with a link to the user-specific keyring. 164 165 When a process changes its real UID, if it used to have no session key, it 166 will be subscribed to the default session key for the new UID. 167 168 If a process attempts to access its session key when it doesn't have one, 169 it will be subscribed to the default for its current UID. 170 171 * Each user has two quotas against which the keys they own are tracked. One 172 limits the total number of keys and keyrings, the other limits the total 173 amount of description and payload space that can be consumed. 174 175 The user can view information on this and other statistics through procfs 176 files. The root user may also alter the quota limits through sysctl files 177 (see the section "New procfs files"). 178 179 Process-specific and thread-specific keyrings are not counted towards a 180 user's quota. 181 182 If a system call that modifies a key or keyring in some way would put the 183 user over quota, the operation is refused and error EDQUOT is returned. 184 185 * There's a system call interface by which userspace programs can create and 186 manipulate keys and keyrings. 187 188 * There's a kernel interface by which services can register types and search 189 for keys. 190 191 * There's a way for the a search done from the kernel to call back to 192 userspace to request a key that can't be found in a process's keyrings. 193 194 * An optional filesystem is available through which the key database can be 195 viewed and manipulated. 196 197 198Key Access Permissions 199====================== 200 201Keys have an owner user ID, a group ID and an ACL. The ACL is made up of a 202sequence of ACEs that each contain three elements: 203 204 * The type of subject. 205 * The subject. 206 207 These two together indicate the subject to whom the permits are granted. 208 The type can be one of: 209 210 * ``KEY_ACE_SUBJ_STANDARD`` 211 212 The subject is a standard 'macro' type. The subject can be one of: 213 214 * ``KEY_ACE_EVERYONE`` 215 216 The permits are granted to everyone. It replaces the old 'other' 217 type on the assumption that you wouldn't grant a permission to other 218 that you you wouldn't grant to everyone else. 219 220 * ``KEY_ACE_OWNER`` 221 222 The permits are granted to the owner of the key (key->uid). 223 224 * ``KEY_ACE_GROUP`` 225 226 The permits are granted to the key's group (key->gid). 227 228 * ``KEY_ACE_POSSESSOR`` 229 230 The permits are granted to anyone who possesses the key. 231 232 * The set of permits granted to the subject. These include: 233 234 * ``KEY_ACE_VIEW`` 235 236 This permits a key or keyring's attributes to be viewed - including the 237 key type and description. 238 239 * ``KEY_ACE_READ`` 240 241 This permits a key's payload to be viewed or a keyring's list of linked 242 keys. 243 244 * ``KEY_ACE_WRITE`` 245 246 This permits a key's payload to be instantiated or updated, or it allows 247 a link to be added to or removed from a keyring. 248 249 * ``KEY_ACE_SEARCH`` 250 251 This permits keyrings to be searched and keys to be found. Searches can 252 only recurse into nested keyrings that have search permission set. 253 254 * ``KEY_ACE_LINK`` 255 256 This permits a key or keyring to be linked to. To create a link from a 257 keyring to a key, a process must have Write permission on the keyring 258 and Link permission on the key. 259 260 * ``KEY_ACE_SET_SECURITY`` 261 262 This permits a key's UID, GID and permissions mask to be changed. 263 264 * ``KEY_ACE_INVAL`` 265 266 This permits a key to be invalidated with KEYCTL_INVALIDATE. 267 268 * ``KEY_ACE_REVOKE`` 269 270 This permits a key to be revoked with KEYCTL_REVOKE. 271 272 * ``KEY_ACE_JOIN`` 273 274 This permits a keyring to be joined as a session by 275 KEYCTL_JOIN_SESSION_KEYRING or KEYCTL_SESSION_TO_PARENT. 276 277 * ``KEY_ACE_CLEAR`` 278 279 This permits a keyring to be cleared. 280 281For changing the ownership, group ID or permissions mask, being the owner of 282the key or having the sysadmin capability is sufficient. 283 284The legacy KEYCTL_SETPERM and KEYCTL_DESCRIBE functions can only see/generate 285View, Read, Write, Search, Link and SetAttr permits, and do this for each of 286possessor, user, group and other permission sets as a 32-bit flag mask. These 287will be approximated/inferred: 288 289 SETPERM Permit Implied ACE Permit 290 =============== ======================= 291 Search Inval, Join 292 Write Revoke, Clear 293 Setattr Set Security, Revoke 294 295 ACE Permit Described as 296 =============== ======================= 297 Inval Search 298 Join Search 299 Revoke Write (unless Setattr) 300 Clear write 301 Set Security Setattr 302 303'Other' will be approximated as/inferred from the 'Everyone' subject. 304 305 306SELinux Support 307=============== 308 309The security class "key" has been added to SELinux so that mandatory access 310controls can be applied to keys created within various contexts. This support 311is preliminary, and is likely to change quite significantly in the near future. 312Currently, all of the basic permissions explained above are provided in SELinux 313as well; SELinux is simply invoked after all basic permission checks have been 314performed. 315 316The value of the file /proc/self/attr/keycreate influences the labeling of 317newly-created keys. If the contents of that file correspond to an SELinux 318security context, then the key will be assigned that context. Otherwise, the 319key will be assigned the current context of the task that invoked the key 320creation request. Tasks must be granted explicit permission to assign a 321particular context to newly-created keys, using the "create" permission in the 322key security class. 323 324The default keyrings associated with users will be labeled with the default 325context of the user if and only if the login programs have been instrumented to 326properly initialize keycreate during the login process. Otherwise, they will 327be labeled with the context of the login program itself. 328 329Note, however, that the default keyrings associated with the root user are 330labeled with the default kernel context, since they are created early in the 331boot process, before root has a chance to log in. 332 333The keyrings associated with new threads are each labeled with the context of 334their associated thread, and both session and process keyrings are handled 335similarly. 336 337 338New ProcFS Files 339================ 340 341Two files have been added to procfs by which an administrator can find out 342about the status of the key service: 343 344 * /proc/keys 345 346 This lists the keys that are currently viewable by the task reading the 347 file, giving information about their type, description and permissions. 348 It is not possible to view the payload of the key this way, though some 349 information about it may be given. 350 351 The only keys included in the list are those that grant View permission to 352 the reading process whether or not it possesses them. Note that LSM 353 security checks are still performed, and may further filter out keys that 354 the current process is not authorised to view. 355 356 The contents of the file look like this:: 357 358 SERIAL FLAGS USAGE EXPY PERM UID GID TYPE DESCRIPTION: SUMMARY 359 00000001 I----- 39 perm 1f3f0000 0 0 keyring _uid_ses.0: 1/4 360 00000002 I----- 2 perm 1f3f0000 0 0 keyring _uid.0: empty 361 00000007 I----- 1 perm 1f3f0000 0 0 keyring _pid.1: empty 362 0000018d I----- 1 perm 1f3f0000 0 0 keyring _pid.412: empty 363 000004d2 I--Q-- 1 perm 1f3f0000 32 -1 keyring _uid.32: 1/4 364 000004d3 I--Q-- 3 perm 1f3f0000 32 -1 keyring _uid_ses.32: empty 365 00000892 I--QU- 1 perm 1f000000 0 0 user metal:copper: 0 366 00000893 I--Q-N 1 35s 1f3f0000 0 0 user metal:silver: 0 367 00000894 I--Q-- 1 10h 003f0000 0 0 user metal:gold: 0 368 369 The flags are:: 370 371 I Instantiated 372 R Revoked 373 D Dead 374 Q Contributes to user's quota 375 U Under construction by callback to userspace 376 N Negative key 377 378 379 * /proc/key-users 380 381 This file lists the tracking data for each user that has at least one key 382 on the system. Such data includes quota information and statistics:: 383 384 [root@andromeda root]# cat /proc/key-users 385 0: 46 45/45 1/100 13/10000 386 29: 2 2/2 2/100 40/10000 387 32: 2 2/2 2/100 40/10000 388 38: 2 2/2 2/100 40/10000 389 390 The format of each line is:: 391 392 <UID>: User ID to which this applies 393 <usage> Structure refcount 394 <inst>/<keys> Total number of keys and number instantiated 395 <keys>/<max> Key count quota 396 <bytes>/<max> Key size quota 397 398 399Four new sysctl files have been added also for the purpose of controlling the 400quota limits on keys: 401 402 * /proc/sys/kernel/keys/root_maxkeys 403 /proc/sys/kernel/keys/root_maxbytes 404 405 These files hold the maximum number of keys that root may have and the 406 maximum total number of bytes of data that root may have stored in those 407 keys. 408 409 * /proc/sys/kernel/keys/maxkeys 410 /proc/sys/kernel/keys/maxbytes 411 412 These files hold the maximum number of keys that each non-root user may 413 have and the maximum total number of bytes of data that each of those 414 users may have stored in their keys. 415 416Root may alter these by writing each new limit as a decimal number string to 417the appropriate file. 418 419 420Userspace System Call Interface 421=============================== 422 423Userspace can manipulate keys directly through three new syscalls: add_key, 424request_key and keyctl. The latter provides a number of functions for 425manipulating keys. 426 427When referring to a key directly, userspace programs should use the key's 428serial number (a positive 32-bit integer). However, there are some special 429values available for referring to special keys and keyrings that relate to the 430process making the call:: 431 432 CONSTANT VALUE KEY REFERENCED 433 ============================== ====== =========================== 434 KEY_SPEC_THREAD_KEYRING -1 thread-specific keyring 435 KEY_SPEC_PROCESS_KEYRING -2 process-specific keyring 436 KEY_SPEC_SESSION_KEYRING -3 session-specific keyring 437 KEY_SPEC_USER_KEYRING -4 UID-specific keyring 438 KEY_SPEC_USER_SESSION_KEYRING -5 UID-session keyring 439 KEY_SPEC_GROUP_KEYRING -6 GID-specific keyring 440 KEY_SPEC_REQKEY_AUTH_KEY -7 assumed request_key() 441 authorisation key 442 443 444The main syscalls are: 445 446 * Create a new key of given type, description and payload and add it to the 447 nominated keyring:: 448 449 key_serial_t add_key(const char *type, const char *desc, 450 const void *payload, size_t plen, 451 key_serial_t keyring); 452 453 If a key of the same type and description as that proposed already exists 454 in the keyring, this will try to update it with the given payload, or it 455 will return error EEXIST if that function is not supported by the key 456 type. The process must also have permission to write to the key to be able 457 to update it. The new key will have all user permissions granted and no 458 group or third party permissions. 459 460 Otherwise, this will attempt to create a new key of the specified type and 461 description, and to instantiate it with the supplied payload and attach it 462 to the keyring. In this case, an error will be generated if the process 463 does not have permission to write to the keyring. 464 465 If the key type supports it, if the description is NULL or an empty 466 string, the key type will try and generate a description from the content 467 of the payload. 468 469 The payload is optional, and the pointer can be NULL if not required by 470 the type. The payload is plen in size, and plen can be zero for an empty 471 payload. 472 473 A new keyring can be generated by setting type "keyring", the keyring name 474 as the description (or NULL) and setting the payload to NULL. 475 476 User defined keys can be created by specifying type "user". It is 477 recommended that a user defined key's description by prefixed with a type 478 ID and a colon, such as "krb5tgt:" for a Kerberos 5 ticket granting 479 ticket. 480 481 Any other type must have been registered with the kernel in advance by a 482 kernel service such as a filesystem. 483 484 The ID of the new or updated key is returned if successful. 485 486 487 * Search the process's keyrings for a key, potentially calling out to 488 userspace to create it:: 489 490 key_serial_t request_key(const char *type, const char *description, 491 const char *callout_info, 492 key_serial_t dest_keyring); 493 494 This function searches all the process's keyrings in the order thread, 495 process, session for a matching key. This works very much like 496 KEYCTL_SEARCH, including the optional attachment of the discovered key to 497 a keyring. 498 499 If a key cannot be found, and if callout_info is not NULL, then 500 /sbin/request-key will be invoked in an attempt to obtain a key. The 501 callout_info string will be passed as an argument to the program. 502 503 To link a key into the destination keyring the key must grant link 504 permission on the key to the caller and the keyring must grant write 505 permission. 506 507 See also Documentation/security/keys/request-key.rst. 508 509 510The keyctl syscall functions are: 511 512 * Map a special key ID to a real key ID for this process:: 513 514 key_serial_t keyctl(KEYCTL_GET_KEYRING_ID, key_serial_t id, 515 int create); 516 517 The special key specified by "id" is looked up (with the key being created 518 if necessary) and the ID of the key or keyring thus found is returned if 519 it exists. 520 521 If the key does not yet exist, the key will be created if "create" is 522 non-zero; and the error ENOKEY will be returned if "create" is zero. 523 524 525 * Replace the session keyring this process subscribes to with a new one:: 526 527 key_serial_t keyctl(KEYCTL_JOIN_SESSION_KEYRING, const char *name); 528 529 If name is NULL, an anonymous keyring is created attached to the process 530 as its session keyring, displacing the old session keyring. 531 532 If name is not NULL, if a keyring of that name exists, the process 533 attempts to attach it as the session keyring, returning an error if that 534 is not permitted; otherwise a new keyring of that name is created and 535 attached as the session keyring. 536 537 To attach to a named keyring, the keyring must have search permission for 538 the process's ownership. 539 540 The ID of the new session keyring is returned if successful. 541 542 543 * Update the specified key:: 544 545 long keyctl(KEYCTL_UPDATE, key_serial_t key, const void *payload, 546 size_t plen); 547 548 This will try to update the specified key with the given payload, or it 549 will return error EOPNOTSUPP if that function is not supported by the key 550 type. The process must also have permission to write to the key to be able 551 to update it. 552 553 The payload is of length plen, and may be absent or empty as for 554 add_key(). 555 556 557 * Revoke a key:: 558 559 long keyctl(KEYCTL_REVOKE, key_serial_t key); 560 561 This makes a key unavailable for further operations. Further attempts to 562 use the key will be met with error EKEYREVOKED, and the key will no longer 563 be findable. 564 565 566 * Change the ownership of a key:: 567 568 long keyctl(KEYCTL_CHOWN, key_serial_t key, uid_t uid, gid_t gid); 569 570 This function permits a key's owner and group ID to be changed. Either one 571 of uid or gid can be set to -1 to suppress that change. 572 573 Only the superuser can change a key's owner to something other than the 574 key's current owner. Similarly, only the superuser can change a key's 575 group ID to something other than the calling process's group ID or one of 576 its group list members. 577 578 579 * Change the permissions mask on a key:: 580 581 long keyctl(KEYCTL_SETPERM, key_serial_t key, key_perm_t perm); 582 583 This function permits the owner of a key or the superuser to change the 584 permissions mask on a key. 585 586 Only bits the available bits are permitted; if any other bits are set, 587 error EINVAL will be returned. 588 589 590 * Describe a key:: 591 592 long keyctl(KEYCTL_DESCRIBE, key_serial_t key, char *buffer, 593 size_t buflen); 594 595 This function returns a summary of the key's attributes (but not its 596 payload data) as a string in the buffer provided. 597 598 Unless there's an error, it always returns the amount of data it could 599 produce, even if that's too big for the buffer, but it won't copy more 600 than requested to userspace. If the buffer pointer is NULL then no copy 601 will take place. 602 603 A process must have view permission on the key for this function to be 604 successful. 605 606 If successful, a string is placed in the buffer in the following format:: 607 608 <type>;<uid>;<gid>;<perm>;<description> 609 610 Where type and description are strings, uid and gid are decimal, and perm 611 is hexadecimal. A NUL character is included at the end of the string if 612 the buffer is sufficiently big. 613 614 This can be parsed with:: 615 616 sscanf(buffer, "%[^;];%d;%d;%o;%s", type, &uid, &gid, &mode, desc); 617 618 619 * Clear out a keyring:: 620 621 long keyctl(KEYCTL_CLEAR, key_serial_t keyring); 622 623 This function clears the list of keys attached to a keyring. The calling 624 process must have write permission on the keyring, and it must be a 625 keyring (or else error ENOTDIR will result). 626 627 This function can also be used to clear special kernel keyrings if they 628 are appropriately marked if the user has CAP_SYS_ADMIN capability. The 629 DNS resolver cache keyring is an example of this. 630 631 632 * Link a key into a keyring:: 633 634 long keyctl(KEYCTL_LINK, key_serial_t keyring, key_serial_t key); 635 636 This function creates a link from the keyring to the key. The process must 637 have write permission on the keyring and must have link permission on the 638 key. 639 640 Should the keyring not be a keyring, error ENOTDIR will result; and if the 641 keyring is full, error ENFILE will result. 642 643 The link procedure checks the nesting of the keyrings, returning ELOOP if 644 it appears too deep or EDEADLK if the link would introduce a cycle. 645 646 Any links within the keyring to keys that match the new key in terms of 647 type and description will be discarded from the keyring as the new one is 648 added. 649 650 651 * Move a key from one keyring to another:: 652 653 long keyctl(KEYCTL_MOVE, 654 key_serial_t id, 655 key_serial_t from_ring_id, 656 key_serial_t to_ring_id, 657 unsigned int flags); 658 659 Move the key specified by "id" from the keyring specified by 660 "from_ring_id" to the keyring specified by "to_ring_id". If the two 661 keyrings are the same, nothing is done. 662 663 "flags" can have KEYCTL_MOVE_EXCL set in it to cause the operation to fail 664 with EEXIST if a matching key exists in the destination keyring, otherwise 665 such a key will be replaced. 666 667 A process must have link permission on the key for this function to be 668 successful and write permission on both keyrings. Any errors that can 669 occur from KEYCTL_LINK also apply on the destination keyring here. 670 671 672 * Unlink a key or keyring from another keyring:: 673 674 long keyctl(KEYCTL_UNLINK, key_serial_t keyring, key_serial_t key); 675 676 This function looks through the keyring for the first link to the 677 specified key, and removes it if found. Subsequent links to that key are 678 ignored. The process must have write permission on the keyring. 679 680 If the keyring is not a keyring, error ENOTDIR will result; and if the key 681 is not present, error ENOENT will be the result. 682 683 684 * Search a keyring tree for a key:: 685 686 key_serial_t keyctl(KEYCTL_SEARCH, key_serial_t keyring, 687 const char *type, const char *description, 688 key_serial_t dest_keyring); 689 690 This searches the keyring tree headed by the specified keyring until a key 691 is found that matches the type and description criteria. Each keyring is 692 checked for keys before recursion into its children occurs. 693 694 The process must have search permission on the top level keyring, or else 695 error EACCES will result. Only keyrings that the process has search 696 permission on will be recursed into, and only keys and keyrings for which 697 a process has search permission can be matched. If the specified keyring 698 is not a keyring, ENOTDIR will result. 699 700 If the search succeeds, the function will attempt to link the found key 701 into the destination keyring if one is supplied (non-zero ID). All the 702 constraints applicable to KEYCTL_LINK apply in this case too. 703 704 Error ENOKEY, EKEYREVOKED or EKEYEXPIRED will be returned if the search 705 fails. On success, the resulting key ID will be returned. 706 707 708 * Read the payload data from a key:: 709 710 long keyctl(KEYCTL_READ, key_serial_t keyring, char *buffer, 711 size_t buflen); 712 713 This function attempts to read the payload data from the specified key 714 into the buffer. The process must have read permission on the key to 715 succeed. 716 717 The returned data will be processed for presentation by the key type. For 718 instance, a keyring will return an array of key_serial_t entries 719 representing the IDs of all the keys to which it is subscribed. The user 720 defined key type will return its data as is. If a key type does not 721 implement this function, error EOPNOTSUPP will result. 722 723 If the specified buffer is too small, then the size of the buffer required 724 will be returned. Note that in this case, the contents of the buffer may 725 have been overwritten in some undefined way. 726 727 Otherwise, on success, the function will return the amount of data copied 728 into the buffer. 729 730 * Instantiate a partially constructed key:: 731 732 long keyctl(KEYCTL_INSTANTIATE, key_serial_t key, 733 const void *payload, size_t plen, 734 key_serial_t keyring); 735 long keyctl(KEYCTL_INSTANTIATE_IOV, key_serial_t key, 736 const struct iovec *payload_iov, unsigned ioc, 737 key_serial_t keyring); 738 739 If the kernel calls back to userspace to complete the instantiation of a 740 key, userspace should use this call to supply data for the key before the 741 invoked process returns, or else the key will be marked negative 742 automatically. 743 744 The process must have write access on the key to be able to instantiate 745 it, and the key must be uninstantiated. 746 747 If a keyring is specified (non-zero), the key will also be linked into 748 that keyring, however all the constraints applying in KEYCTL_LINK apply in 749 this case too. 750 751 The payload and plen arguments describe the payload data as for add_key(). 752 753 The payload_iov and ioc arguments describe the payload data in an iovec 754 array instead of a single buffer. 755 756 757 * Negatively instantiate a partially constructed key:: 758 759 long keyctl(KEYCTL_NEGATE, key_serial_t key, 760 unsigned timeout, key_serial_t keyring); 761 long keyctl(KEYCTL_REJECT, key_serial_t key, 762 unsigned timeout, unsigned error, key_serial_t keyring); 763 764 If the kernel calls back to userspace to complete the instantiation of a 765 key, userspace should use this call mark the key as negative before the 766 invoked process returns if it is unable to fulfill the request. 767 768 The process must have write access on the key to be able to instantiate 769 it, and the key must be uninstantiated. 770 771 If a keyring is specified (non-zero), the key will also be linked into 772 that keyring, however all the constraints applying in KEYCTL_LINK apply in 773 this case too. 774 775 If the key is rejected, future searches for it will return the specified 776 error code until the rejected key expires. Negating the key is the same 777 as rejecting the key with ENOKEY as the error code. 778 779 780 * Set the default request-key destination keyring:: 781 782 long keyctl(KEYCTL_SET_REQKEY_KEYRING, int reqkey_defl); 783 784 This sets the default keyring to which implicitly requested keys will be 785 attached for this thread. reqkey_defl should be one of these constants:: 786 787 CONSTANT VALUE NEW DEFAULT KEYRING 788 ====================================== ====== ======================= 789 KEY_REQKEY_DEFL_NO_CHANGE -1 No change 790 KEY_REQKEY_DEFL_DEFAULT 0 Default[1] 791 KEY_REQKEY_DEFL_THREAD_KEYRING 1 Thread keyring 792 KEY_REQKEY_DEFL_PROCESS_KEYRING 2 Process keyring 793 KEY_REQKEY_DEFL_SESSION_KEYRING 3 Session keyring 794 KEY_REQKEY_DEFL_USER_KEYRING 4 User keyring 795 KEY_REQKEY_DEFL_USER_SESSION_KEYRING 5 User session keyring 796 KEY_REQKEY_DEFL_GROUP_KEYRING 6 Group keyring 797 798 The old default will be returned if successful and error EINVAL will be 799 returned if reqkey_defl is not one of the above values. 800 801 The default keyring can be overridden by the keyring indicated to the 802 request_key() system call. 803 804 Note that this setting is inherited across fork/exec. 805 806 [1] The default is: the thread keyring if there is one, otherwise 807 the process keyring if there is one, otherwise the session keyring if 808 there is one, otherwise the user default session keyring. 809 810 811 * Set the timeout on a key:: 812 813 long keyctl(KEYCTL_SET_TIMEOUT, key_serial_t key, unsigned timeout); 814 815 This sets or clears the timeout on a key. The timeout can be 0 to clear 816 the timeout or a number of seconds to set the expiry time that far into 817 the future. 818 819 The process must have attribute modification access on a key to set its 820 timeout. Timeouts may not be set with this function on negative, revoked 821 or expired keys. 822 823 824 * Assume the authority granted to instantiate a key:: 825 826 long keyctl(KEYCTL_ASSUME_AUTHORITY, key_serial_t key); 827 828 This assumes or divests the authority required to instantiate the 829 specified key. Authority can only be assumed if the thread has the 830 authorisation key associated with the specified key in its keyrings 831 somewhere. 832 833 Once authority is assumed, searches for keys will also search the 834 requester's keyrings using the requester's security label, UID, GID and 835 groups. 836 837 If the requested authority is unavailable, error EPERM will be returned, 838 likewise if the authority has been revoked because the target key is 839 already instantiated. 840 841 If the specified key is 0, then any assumed authority will be divested. 842 843 The assumed authoritative key is inherited across fork and exec. 844 845 846 * Get the LSM security context attached to a key:: 847 848 long keyctl(KEYCTL_GET_SECURITY, key_serial_t key, char *buffer, 849 size_t buflen) 850 851 This function returns a string that represents the LSM security context 852 attached to a key in the buffer provided. 853 854 Unless there's an error, it always returns the amount of data it could 855 produce, even if that's too big for the buffer, but it won't copy more 856 than requested to userspace. If the buffer pointer is NULL then no copy 857 will take place. 858 859 A NUL character is included at the end of the string if the buffer is 860 sufficiently big. This is included in the returned count. If no LSM is 861 in force then an empty string will be returned. 862 863 A process must have view permission on the key for this function to be 864 successful. 865 866 867 * Install the calling process's session keyring on its parent:: 868 869 long keyctl(KEYCTL_SESSION_TO_PARENT); 870 871 This functions attempts to install the calling process's session keyring 872 on to the calling process's parent, replacing the parent's current session 873 keyring. 874 875 The calling process must have the same ownership as its parent, the 876 keyring must have the same ownership as the calling process, the calling 877 process must have LINK permission on the keyring and the active LSM module 878 mustn't deny permission, otherwise error EPERM will be returned. 879 880 Error ENOMEM will be returned if there was insufficient memory to complete 881 the operation, otherwise 0 will be returned to indicate success. 882 883 The keyring will be replaced next time the parent process leaves the 884 kernel and resumes executing userspace. 885 886 887 * Invalidate a key:: 888 889 long keyctl(KEYCTL_INVALIDATE, key_serial_t key); 890 891 This function marks a key as being invalidated and then wakes up the 892 garbage collector. The garbage collector immediately removes invalidated 893 keys from all keyrings and deletes the key when its reference count 894 reaches zero. 895 896 Keys that are marked invalidated become invisible to normal key operations 897 immediately, though they are still visible in /proc/keys until deleted 898 (they're marked with an 'i' flag). 899 900 A process must have search permission on the key for this function to be 901 successful. 902 903 * Compute a Diffie-Hellman shared secret or public key:: 904 905 long keyctl(KEYCTL_DH_COMPUTE, struct keyctl_dh_params *params, 906 char *buffer, size_t buflen, struct keyctl_kdf_params *kdf); 907 908 The params struct contains serial numbers for three keys:: 909 910 - The prime, p, known to both parties 911 - The local private key 912 - The base integer, which is either a shared generator or the 913 remote public key 914 915 The value computed is:: 916 917 result = base ^ private (mod prime) 918 919 If the base is the shared generator, the result is the local 920 public key. If the base is the remote public key, the result is 921 the shared secret. 922 923 If the parameter kdf is NULL, the following applies: 924 925 - The buffer length must be at least the length of the prime, or zero. 926 927 - If the buffer length is nonzero, the length of the result is 928 returned when it is successfully calculated and copied in to the 929 buffer. When the buffer length is zero, the minimum required 930 buffer length is returned. 931 932 The kdf parameter allows the caller to apply a key derivation function 933 (KDF) on the Diffie-Hellman computation where only the result 934 of the KDF is returned to the caller. The KDF is characterized with 935 struct keyctl_kdf_params as follows: 936 937 - ``char *hashname`` specifies the NUL terminated string identifying 938 the hash used from the kernel crypto API and applied for the KDF 939 operation. The KDF implemenation complies with SP800-56A as well 940 as with SP800-108 (the counter KDF). 941 942 - ``char *otherinfo`` specifies the OtherInfo data as documented in 943 SP800-56A section 5.8.1.2. The length of the buffer is given with 944 otherinfolen. The format of OtherInfo is defined by the caller. 945 The otherinfo pointer may be NULL if no OtherInfo shall be used. 946 947 This function will return error EOPNOTSUPP if the key type is not 948 supported, error ENOKEY if the key could not be found, or error 949 EACCES if the key is not readable by the caller. In addition, the 950 function will return EMSGSIZE when the parameter kdf is non-NULL 951 and either the buffer length or the OtherInfo length exceeds the 952 allowed length. 953 954 955 * Restrict keyring linkage:: 956 957 long keyctl(KEYCTL_RESTRICT_KEYRING, key_serial_t keyring, 958 const char *type, const char *restriction); 959 960 An existing keyring can restrict linkage of additional keys by evaluating 961 the contents of the key according to a restriction scheme. 962 963 "keyring" is the key ID for an existing keyring to apply a restriction 964 to. It may be empty or may already have keys linked. Existing linked keys 965 will remain in the keyring even if the new restriction would reject them. 966 967 "type" is a registered key type. 968 969 "restriction" is a string describing how key linkage is to be restricted. 970 The format varies depending on the key type, and the string is passed to 971 the lookup_restriction() function for the requested type. It may specify 972 a method and relevant data for the restriction such as signature 973 verification or constraints on key payload. If the requested key type is 974 later unregistered, no keys may be added to the keyring after the key type 975 is removed. 976 977 To apply a keyring restriction the process must have Set Attribute 978 permission and the keyring must not be previously restricted. 979 980 One application of restricted keyrings is to verify X.509 certificate 981 chains or individual certificate signatures using the asymmetric key type. 982 See Documentation/crypto/asymmetric-keys.txt for specific restrictions 983 applicable to the asymmetric key type. 984 985 986 * Query an asymmetric key:: 987 988 long keyctl(KEYCTL_PKEY_QUERY, 989 key_serial_t key_id, unsigned long reserved, 990 struct keyctl_pkey_query *info); 991 992 Get information about an asymmetric key. The information is returned in 993 the keyctl_pkey_query struct:: 994 995 __u32 supported_ops; 996 __u32 key_size; 997 __u16 max_data_size; 998 __u16 max_sig_size; 999 __u16 max_enc_size; 1000 __u16 max_dec_size; 1001 __u32 __spare[10]; 1002 1003 ``supported_ops`` contains a bit mask of flags indicating which ops are 1004 supported. This is constructed from a bitwise-OR of:: 1005 1006 KEYCTL_SUPPORTS_{ENCRYPT,DECRYPT,SIGN,VERIFY} 1007 1008 ``key_size`` indicated the size of the key in bits. 1009 1010 ``max_*_size`` indicate the maximum sizes in bytes of a blob of data to be 1011 signed, a signature blob, a blob to be encrypted and a blob to be 1012 decrypted. 1013 1014 ``__spare[]`` must be set to 0. This is intended for future use to hand 1015 over one or more passphrases needed unlock a key. 1016 1017 If successful, 0 is returned. If the key is not an asymmetric key, 1018 EOPNOTSUPP is returned. 1019 1020 1021 * Encrypt, decrypt, sign or verify a blob using an asymmetric key:: 1022 1023 long keyctl(KEYCTL_PKEY_ENCRYPT, 1024 const struct keyctl_pkey_params *params, 1025 const char *info, 1026 const void *in, 1027 void *out); 1028 1029 long keyctl(KEYCTL_PKEY_DECRYPT, 1030 const struct keyctl_pkey_params *params, 1031 const char *info, 1032 const void *in, 1033 void *out); 1034 1035 long keyctl(KEYCTL_PKEY_SIGN, 1036 const struct keyctl_pkey_params *params, 1037 const char *info, 1038 const void *in, 1039 void *out); 1040 1041 long keyctl(KEYCTL_PKEY_VERIFY, 1042 const struct keyctl_pkey_params *params, 1043 const char *info, 1044 const void *in, 1045 const void *in2); 1046 1047 Use an asymmetric key to perform a public-key cryptographic operation a 1048 blob of data. For encryption and verification, the asymmetric key may 1049 only need the public parts to be available, but for decryption and signing 1050 the private parts are required also. 1051 1052 The parameter block pointed to by params contains a number of integer 1053 values:: 1054 1055 __s32 key_id; 1056 __u32 in_len; 1057 __u32 out_len; 1058 __u32 in2_len; 1059 1060 ``key_id`` is the ID of the asymmetric key to be used. ``in_len`` and 1061 ``in2_len`` indicate the amount of data in the in and in2 buffers and 1062 ``out_len`` indicates the size of the out buffer as appropriate for the 1063 above operations. 1064 1065 For a given operation, the in and out buffers are used as follows:: 1066 1067 Operation ID in,in_len out,out_len in2,in2_len 1068 ======================= =============== =============== =============== 1069 KEYCTL_PKEY_ENCRYPT Raw data Encrypted data - 1070 KEYCTL_PKEY_DECRYPT Encrypted data Raw data - 1071 KEYCTL_PKEY_SIGN Raw data Signature - 1072 KEYCTL_PKEY_VERIFY Raw data - Signature 1073 1074 ``info`` is a string of key=value pairs that supply supplementary 1075 information. These include: 1076 1077 ``enc=<encoding>`` The encoding of the encrypted/signature blob. This 1078 can be "pkcs1" for RSASSA-PKCS1-v1.5 or 1079 RSAES-PKCS1-v1.5; "pss" for "RSASSA-PSS"; "oaep" for 1080 "RSAES-OAEP". If omitted or is "raw", the raw output 1081 of the encryption function is specified. 1082 1083 ``hash=<algo>`` If the data buffer contains the output of a hash 1084 function and the encoding includes some indication of 1085 which hash function was used, the hash function can be 1086 specified with this, eg. "hash=sha256". 1087 1088 The ``__spare[]`` space in the parameter block must be set to 0. This is 1089 intended, amongst other things, to allow the passing of passphrases 1090 required to unlock a key. 1091 1092 If successful, encrypt, decrypt and sign all return the amount of data 1093 written into the output buffer. Verification returns 0 on success. 1094 1095 1096Kernel Services 1097=============== 1098 1099The kernel services for key management are fairly simple to deal with. They can 1100be broken down into two areas: keys and key types. 1101 1102Dealing with keys is fairly straightforward. Firstly, the kernel service 1103registers its type, then it searches for a key of that type. It should retain 1104the key as long as it has need of it, and then it should release it. For a 1105filesystem or device file, a search would probably be performed during the open 1106call, and the key released upon close. How to deal with conflicting keys due to 1107two different users opening the same file is left to the filesystem author to 1108solve. 1109 1110To access the key manager, the following header must be #included:: 1111 1112 <linux/key.h> 1113 1114Specific key types should have a header file under include/keys/ that should be 1115used to access that type. For keys of type "user", for example, that would be:: 1116 1117 <keys/user-type.h> 1118 1119Note that there are two different types of pointers to keys that may be 1120encountered: 1121 1122 * struct key * 1123 1124 This simply points to the key structure itself. Key structures will be at 1125 least four-byte aligned. 1126 1127 * key_ref_t 1128 1129 This is equivalent to a ``struct key *``, but the least significant bit is set 1130 if the caller "possesses" the key. By "possession" it is meant that the 1131 calling processes has a searchable link to the key from one of its 1132 keyrings. There are three functions for dealing with these:: 1133 1134 key_ref_t make_key_ref(const struct key *key, bool possession); 1135 1136 struct key *key_ref_to_ptr(const key_ref_t key_ref); 1137 1138 bool is_key_possessed(const key_ref_t key_ref); 1139 1140 The first function constructs a key reference from a key pointer and 1141 possession information (which must be true or false). 1142 1143 The second function retrieves the key pointer from a reference and the 1144 third retrieves the possession flag. 1145 1146When accessing a key's payload contents, certain precautions must be taken to 1147prevent access vs modification races. See the section "Notes on accessing 1148payload contents" for more information. 1149 1150 * To search for a key, call:: 1151 1152 struct key *request_key(const struct key_type *type, 1153 const char *description, 1154 const char *callout_info, 1155 struct key_acl *acl); 1156 1157 This is used to request a key or keyring with a description that matches 1158 the description specified according to the key type's match_preparse() 1159 method. This permits approximate matching to occur. If callout_string is 1160 not NULL, then /sbin/request-key will be invoked in an attempt to obtain 1161 the key from userspace. In that case, callout_string will be passed as an 1162 argument to the program. 1163 1164 Should the function fail error ENOKEY, EKEYEXPIRED or EKEYREVOKED will be 1165 returned. 1166 1167 If successful, the key will have been attached to the default keyring for 1168 implicitly obtained request-key keys, as set by KEYCTL_SET_REQKEY_KEYRING. 1169 1170 If a key is created, it will be given the specified ACL. 1171 1172 See also Documentation/security/keys/request-key.rst. 1173 1174 1175 * To search for a key in a specific domain, call: 1176 1177 struct key *request_key_tag(const struct key_type *type, 1178 const char *description, 1179 struct key_tag *domain_tag, 1180 const char *callout_info, 1181 struct key_acl *acl); 1182 1183 This is identical to request_key(), except that a domain tag may be 1184 specifies that causes search algorithm to only match keys matching that 1185 tag. The domain_tag may be NULL, specifying a global domain that is 1186 separate from any nominated domain. 1187 1188 1189 * To search for a key, passing auxiliary data to the upcaller, call:: 1190 1191 struct key *request_key_with_auxdata(const struct key_type *type, 1192 const char *description, 1193 struct key_tag *domain_tag, 1194 const void *callout_info, 1195 size_t callout_len, 1196 void *aux, 1197 struct key_acl *acl); 1198 1199 This is identical to request_key_tag(), except that the auxiliary data is 1200 passed to the key_type->request_key() op if it exists, and the 1201 callout_info is a blob of length callout_len, if given (the length may be 1202 0). 1203 1204 1205 * To search for a key under RCU conditions, call:: 1206 1207 struct key *request_key_rcu(const struct key_type *type, 1208 const char *description, 1209 struct key_tag *domain_tag); 1210 1211 which is similar to request_key_tag() except that it does not check for 1212 keys that are under construction and it will not call out to userspace to 1213 construct a key if it can't find a match. 1214 1215 1216 * When it is no longer required, the key should be released using:: 1217 1218 void key_put(struct key *key); 1219 1220 Or:: 1221 1222 void key_ref_put(key_ref_t key_ref); 1223 1224 These can be called from interrupt context. If CONFIG_KEYS is not set then 1225 the argument will not be parsed. 1226 1227 1228 * Extra references can be made to a key by calling one of the following 1229 functions:: 1230 1231 struct key *__key_get(struct key *key); 1232 struct key *key_get(struct key *key); 1233 1234 Keys so references will need to be disposed of by calling key_put() when 1235 they've been finished with. The key pointer passed in will be returned. 1236 1237 In the case of key_get(), if the pointer is NULL or CONFIG_KEYS is not set 1238 then the key will not be dereferenced and no increment will take place. 1239 1240 1241 * A key's serial number can be obtained by calling:: 1242 1243 key_serial_t key_serial(struct key *key); 1244 1245 If key is NULL or if CONFIG_KEYS is not set then 0 will be returned (in the 1246 latter case without parsing the argument). 1247 1248 1249 * If a keyring was found in the search, this can be further searched by:: 1250 1251 key_ref_t keyring_search(key_ref_t keyring_ref, 1252 const struct key_type *type, 1253 const char *description, 1254 bool recurse) 1255 1256 This searches the specified keyring only (recurse == false) or keyring tree 1257 (recurse == true) specified for a matching key. Error ENOKEY is returned 1258 upon failure (use IS_ERR/PTR_ERR to determine). If successful, the returned 1259 key will need to be released. 1260 1261 The possession attribute from the keyring reference is used to control 1262 access through the permissions mask and is propagated to the returned key 1263 reference pointer if successful. 1264 1265 1266 * A keyring can be created by:: 1267 1268 struct key *keyring_alloc(const char *description, uid_t uid, gid_t gid, 1269 const struct cred *cred, 1270 struct key_acl *acl, 1271 struct key_restriction *restrict_link, 1272 unsigned long flags, 1273 struct key *dest); 1274 1275 This creates a keyring with the given attributes and returns it. If dest 1276 is not NULL, the new keyring will be linked into the keyring to which it 1277 points. No permission checks are made upon the destination keyring. 1278 1279 Error EDQUOT can be returned if the keyring would overload the quota (pass 1280 KEY_ALLOC_NOT_IN_QUOTA in flags if the keyring shouldn't be accounted 1281 towards the user's quota). Error ENOMEM can also be returned. 1282 1283 If restrict_link is not NULL, it should point to a structure that contains 1284 the function that will be called each time an attempt is made to link a 1285 key into the new keyring. The structure may also contain a key pointer 1286 and an associated key type. The function is called to check whether a key 1287 may be added into the keyring or not. The key type is used by the garbage 1288 collector to clean up function or data pointers in this structure if the 1289 given key type is unregistered. Callers of key_create_or_update() within 1290 the kernel can pass KEY_ALLOC_BYPASS_RESTRICTION to suppress the check. 1291 An example of using this is to manage rings of cryptographic keys that are 1292 set up when the kernel boots where userspace is also permitted to add keys 1293 - provided they can be verified by a key the kernel already has. 1294 1295 When called, the restriction function will be passed the keyring being 1296 added to, the key type, the payload of the key being added, and data to be 1297 used in the restriction check. Note that when a new key is being created, 1298 this is called between payload preparsing and actual key creation. The 1299 function should return 0 to allow the link or an error to reject it. 1300 1301 A convenience function, restrict_link_reject, exists to always return 1302 -EPERM to in this case. 1303 1304 1305 * To check the validity of a key, this function can be called:: 1306 1307 int validate_key(struct key *key); 1308 1309 This checks that the key in question hasn't expired or and hasn't been 1310 revoked. Should the key be invalid, error EKEYEXPIRED or EKEYREVOKED will 1311 be returned. If the key is NULL or if CONFIG_KEYS is not set then 0 will be 1312 returned (in the latter case without parsing the argument). 1313 1314 1315 * To register a key type, the following function should be called:: 1316 1317 int register_key_type(struct key_type *type); 1318 1319 This will return error EEXIST if a type of the same name is already 1320 present. 1321 1322 1323 * To unregister a key type, call:: 1324 1325 void unregister_key_type(struct key_type *type); 1326 1327 1328Under some circumstances, it may be desirable to deal with a bundle of keys. 1329The facility provides access to the keyring type for managing such a bundle:: 1330 1331 struct key_type key_type_keyring; 1332 1333This can be used with a function such as request_key() to find a specific 1334keyring in a process's keyrings. A keyring thus found can then be searched 1335with keyring_search(). Note that it is not possible to use request_key() to 1336search a specific keyring, so using keyrings in this way is of limited utility. 1337 1338 1339Notes On Accessing Payload Contents 1340=================================== 1341 1342The simplest payload is just data stored in key->payload directly. In this 1343case, there's no need to indulge in RCU or locking when accessing the payload. 1344 1345More complex payload contents must be allocated and pointers to them set in the 1346key->payload.data[] array. One of the following ways must be selected to 1347access the data: 1348 1349 1) Unmodifiable key type. 1350 1351 If the key type does not have a modify method, then the key's payload can 1352 be accessed without any form of locking, provided that it's known to be 1353 instantiated (uninstantiated keys cannot be "found"). 1354 1355 2) The key's semaphore. 1356 1357 The semaphore could be used to govern access to the payload and to control 1358 the payload pointer. It must be write-locked for modifications and would 1359 have to be read-locked for general access. The disadvantage of doing this 1360 is that the accessor may be required to sleep. 1361 1362 3) RCU. 1363 1364 RCU must be used when the semaphore isn't already held; if the semaphore 1365 is held then the contents can't change under you unexpectedly as the 1366 semaphore must still be used to serialise modifications to the key. The 1367 key management code takes care of this for the key type. 1368 1369 However, this means using:: 1370 1371 rcu_read_lock() ... rcu_dereference() ... rcu_read_unlock() 1372 1373 to read the pointer, and:: 1374 1375 rcu_dereference() ... rcu_assign_pointer() ... call_rcu() 1376 1377 to set the pointer and dispose of the old contents after a grace period. 1378 Note that only the key type should ever modify a key's payload. 1379 1380 Furthermore, an RCU controlled payload must hold a struct rcu_head for the 1381 use of call_rcu() and, if the payload is of variable size, the length of 1382 the payload. key->datalen cannot be relied upon to be consistent with the 1383 payload just dereferenced if the key's semaphore is not held. 1384 1385 Note that key->payload.data[0] has a shadow that is marked for __rcu 1386 usage. This is called key->payload.rcu_data0. The following accessors 1387 wrap the RCU calls to this element: 1388 1389 a) Set or change the first payload pointer:: 1390 1391 rcu_assign_keypointer(struct key *key, void *data); 1392 1393 b) Read the first payload pointer with the key semaphore held:: 1394 1395 [const] void *dereference_key_locked([const] struct key *key); 1396 1397 Note that the return value will inherit its constness from the key 1398 parameter. Static analysis will give an error if it things the lock 1399 isn't held. 1400 1401 c) Read the first payload pointer with the RCU read lock held:: 1402 1403 const void *dereference_key_rcu(const struct key *key); 1404 1405 1406Defining a Key Type 1407=================== 1408 1409A kernel service may want to define its own key type. For instance, an AFS 1410filesystem might want to define a Kerberos 5 ticket key type. To do this, it 1411author fills in a key_type struct and registers it with the system. 1412 1413Source files that implement key types should include the following header file:: 1414 1415 <linux/key-type.h> 1416 1417The structure has a number of fields, some of which are mandatory: 1418 1419 * ``const char *name`` 1420 1421 The name of the key type. This is used to translate a key type name 1422 supplied by userspace into a pointer to the structure. 1423 1424 1425 * ``size_t def_datalen`` 1426 1427 This is optional - it supplies the default payload data length as 1428 contributed to the quota. If the key type's payload is always or almost 1429 always the same size, then this is a more efficient way to do things. 1430 1431 The data length (and quota) on a particular key can always be changed 1432 during instantiation or update by calling:: 1433 1434 int key_payload_reserve(struct key *key, size_t datalen); 1435 1436 With the revised data length. Error EDQUOT will be returned if this is not 1437 viable. 1438 1439 1440 * ``int (*vet_description)(const char *description);`` 1441 1442 This optional method is called to vet a key description. If the key type 1443 doesn't approve of the key description, it may return an error, otherwise 1444 it should return 0. 1445 1446 1447 * ``int (*preparse)(struct key_preparsed_payload *prep);`` 1448 1449 This optional method permits the key type to attempt to parse payload 1450 before a key is created (add key) or the key semaphore is taken (update or 1451 instantiate key). The structure pointed to by prep looks like:: 1452 1453 struct key_preparsed_payload { 1454 char *description; 1455 union key_payload payload; 1456 const void *data; 1457 size_t datalen; 1458 size_t quotalen; 1459 time_t expiry; 1460 }; 1461 1462 Before calling the method, the caller will fill in data and datalen with 1463 the payload blob parameters; quotalen will be filled in with the default 1464 quota size from the key type; expiry will be set to TIME_T_MAX and the 1465 rest will be cleared. 1466 1467 If a description can be proposed from the payload contents, that should be 1468 attached as a string to the description field. This will be used for the 1469 key description if the caller of add_key() passes NULL or "". 1470 1471 The method can attach anything it likes to payload. This is merely passed 1472 along to the instantiate() or update() operations. If set, the expiry 1473 time will be applied to the key if it is instantiated from this data. 1474 1475 The method should return 0 if successful or a negative error code 1476 otherwise. 1477 1478 1479 * ``void (*free_preparse)(struct key_preparsed_payload *prep);`` 1480 1481 This method is only required if the preparse() method is provided, 1482 otherwise it is unused. It cleans up anything attached to the description 1483 and payload fields of the key_preparsed_payload struct as filled in by the 1484 preparse() method. It will always be called after preparse() returns 1485 successfully, even if instantiate() or update() succeed. 1486 1487 1488 * ``int (*instantiate)(struct key *key, struct key_preparsed_payload *prep);`` 1489 1490 This method is called to attach a payload to a key during construction. 1491 The payload attached need not bear any relation to the data passed to this 1492 function. 1493 1494 The prep->data and prep->datalen fields will define the original payload 1495 blob. If preparse() was supplied then other fields may be filled in also. 1496 1497 If the amount of data attached to the key differs from the size in 1498 keytype->def_datalen, then key_payload_reserve() should be called. 1499 1500 This method does not have to lock the key in order to attach a payload. 1501 The fact that KEY_FLAG_INSTANTIATED is not set in key->flags prevents 1502 anything else from gaining access to the key. 1503 1504 It is safe to sleep in this method. 1505 1506 generic_key_instantiate() is provided to simply copy the data from 1507 prep->payload.data[] to key->payload.data[], with RCU-safe assignment on 1508 the first element. It will then clear prep->payload.data[] so that the 1509 free_preparse method doesn't release the data. 1510 1511 1512 * ``int (*update)(struct key *key, const void *data, size_t datalen);`` 1513 1514 If this type of key can be updated, then this method should be provided. 1515 It is called to update a key's payload from the blob of data provided. 1516 1517 The prep->data and prep->datalen fields will define the original payload 1518 blob. If preparse() was supplied then other fields may be filled in also. 1519 1520 key_payload_reserve() should be called if the data length might change 1521 before any changes are actually made. Note that if this succeeds, the type 1522 is committed to changing the key because it's already been altered, so all 1523 memory allocation must be done first. 1524 1525 The key will have its semaphore write-locked before this method is called, 1526 but this only deters other writers; any changes to the key's payload must 1527 be made under RCU conditions, and call_rcu() must be used to dispose of 1528 the old payload. 1529 1530 key_payload_reserve() should be called before the changes are made, but 1531 after all allocations and other potentially failing function calls are 1532 made. 1533 1534 It is safe to sleep in this method. 1535 1536 1537 * ``int (*match_preparse)(struct key_match_data *match_data);`` 1538 1539 This method is optional. It is called when a key search is about to be 1540 performed. It is given the following structure:: 1541 1542 struct key_match_data { 1543 bool (*cmp)(const struct key *key, 1544 const struct key_match_data *match_data); 1545 const void *raw_data; 1546 void *preparsed; 1547 unsigned lookup_type; 1548 }; 1549 1550 On entry, raw_data will be pointing to the criteria to be used in matching 1551 a key by the caller and should not be modified. ``(*cmp)()`` will be pointing 1552 to the default matcher function (which does an exact description match 1553 against raw_data) and lookup_type will be set to indicate a direct lookup. 1554 1555 The following lookup_type values are available: 1556 1557 * KEYRING_SEARCH_LOOKUP_DIRECT - A direct lookup hashes the type and 1558 description to narrow down the search to a small number of keys. 1559 1560 * KEYRING_SEARCH_LOOKUP_ITERATE - An iterative lookup walks all the 1561 keys in the keyring until one is matched. This must be used for any 1562 search that's not doing a simple direct match on the key description. 1563 1564 The method may set cmp to point to a function of its choice that does some 1565 other form of match, may set lookup_type to KEYRING_SEARCH_LOOKUP_ITERATE 1566 and may attach something to the preparsed pointer for use by ``(*cmp)()``. 1567 ``(*cmp)()`` should return true if a key matches and false otherwise. 1568 1569 If preparsed is set, it may be necessary to use the match_free() method to 1570 clean it up. 1571 1572 The method should return 0 if successful or a negative error code 1573 otherwise. 1574 1575 It is permitted to sleep in this method, but ``(*cmp)()`` may not sleep as 1576 locks will be held over it. 1577 1578 If match_preparse() is not provided, keys of this type will be matched 1579 exactly by their description. 1580 1581 1582 * ``void (*match_free)(struct key_match_data *match_data);`` 1583 1584 This method is optional. If given, it called to clean up 1585 match_data->preparsed after a successful call to match_preparse(). 1586 1587 1588 * ``void (*revoke)(struct key *key);`` 1589 1590 This method is optional. It is called to discard part of the payload 1591 data upon a key being revoked. The caller will have the key semaphore 1592 write-locked. 1593 1594 It is safe to sleep in this method, though care should be taken to avoid 1595 a deadlock against the key semaphore. 1596 1597 1598 * ``void (*destroy)(struct key *key);`` 1599 1600 This method is optional. It is called to discard the payload data on a key 1601 when it is being destroyed. 1602 1603 This method does not need to lock the key to access the payload; it can 1604 consider the key as being inaccessible at this time. Note that the key's 1605 type may have been changed before this function is called. 1606 1607 It is not safe to sleep in this method; the caller may hold spinlocks. 1608 1609 1610 * ``void (*describe)(const struct key *key, struct seq_file *p);`` 1611 1612 This method is optional. It is called during /proc/keys reading to 1613 summarise a key's description and payload in text form. 1614 1615 This method will be called with the RCU read lock held. rcu_dereference() 1616 should be used to read the payload pointer if the payload is to be 1617 accessed. key->datalen cannot be trusted to stay consistent with the 1618 contents of the payload. 1619 1620 The description will not change, though the key's state may. 1621 1622 It is not safe to sleep in this method; the RCU read lock is held by the 1623 caller. 1624 1625 1626 * ``long (*read)(const struct key *key, char __user *buffer, size_t buflen);`` 1627 1628 This method is optional. It is called by KEYCTL_READ to translate the 1629 key's payload into something a blob of data for userspace to deal with. 1630 Ideally, the blob should be in the same format as that passed in to the 1631 instantiate and update methods. 1632 1633 If successful, the blob size that could be produced should be returned 1634 rather than the size copied. 1635 1636 This method will be called with the key's semaphore read-locked. This will 1637 prevent the key's payload changing. It is not necessary to use RCU locking 1638 when accessing the key's payload. It is safe to sleep in this method, such 1639 as might happen when the userspace buffer is accessed. 1640 1641 1642 * ``int (*request_key)(struct key_construction *cons, const char *op, void *aux);`` 1643 1644 This method is optional. If provided, request_key() and friends will 1645 invoke this function rather than upcalling to /sbin/request-key to operate 1646 upon a key of this type. 1647 1648 The aux parameter is as passed to request_key_async_with_auxdata() and 1649 similar or is NULL otherwise. Also passed are the construction record for 1650 the key to be operated upon and the operation type (currently only 1651 "create"). 1652 1653 This method is permitted to return before the upcall is complete, but the 1654 following function must be called under all circumstances to complete the 1655 instantiation process, whether or not it succeeds, whether or not there's 1656 an error:: 1657 1658 void complete_request_key(struct key_construction *cons, int error); 1659 1660 The error parameter should be 0 on success, -ve on error. The 1661 construction record is destroyed by this action and the authorisation key 1662 will be revoked. If an error is indicated, the key under construction 1663 will be negatively instantiated if it wasn't already instantiated. 1664 1665 If this method returns an error, that error will be returned to the 1666 caller of request_key*(). complete_request_key() must be called prior to 1667 returning. 1668 1669 The key under construction and the authorisation key can be found in the 1670 key_construction struct pointed to by cons: 1671 1672 * ``struct key *key;`` 1673 1674 The key under construction. 1675 1676 * ``struct key *authkey;`` 1677 1678 The authorisation key. 1679 1680 1681 * ``struct key_restriction *(*lookup_restriction)(const char *params);`` 1682 1683 This optional method is used to enable userspace configuration of keyring 1684 restrictions. The restriction parameter string (not including the key type 1685 name) is passed in, and this method returns a pointer to a key_restriction 1686 structure containing the relevant functions and data to evaluate each 1687 attempted key link operation. If there is no match, -EINVAL is returned. 1688 1689 1690 * ``int (*asym_eds_op)(struct kernel_pkey_params *params, 1691 const void *in, void *out);`` 1692 ``int (*asym_verify_signature)(struct kernel_pkey_params *params, 1693 const void *in, const void *in2);`` 1694 1695 These methods are optional. If provided the first allows a key to be 1696 used to encrypt, decrypt or sign a blob of data, and the second allows a 1697 key to verify a signature. 1698 1699 In all cases, the following information is provided in the params block:: 1700 1701 struct kernel_pkey_params { 1702 struct key *key; 1703 const char *encoding; 1704 const char *hash_algo; 1705 char *info; 1706 __u32 in_len; 1707 union { 1708 __u32 out_len; 1709 __u32 in2_len; 1710 }; 1711 enum kernel_pkey_operation op : 8; 1712 }; 1713 1714 This includes the key to be used; a string indicating the encoding to use 1715 (for instance, "pkcs1" may be used with an RSA key to indicate 1716 RSASSA-PKCS1-v1.5 or RSAES-PKCS1-v1.5 encoding or "raw" if no encoding); 1717 the name of the hash algorithm used to generate the data for a signature 1718 (if appropriate); the sizes of the input and output (or second input) 1719 buffers; and the ID of the operation to be performed. 1720 1721 For a given operation ID, the input and output buffers are used as 1722 follows:: 1723 1724 Operation ID in,in_len out,out_len in2,in2_len 1725 ======================= =============== =============== =============== 1726 kernel_pkey_encrypt Raw data Encrypted data - 1727 kernel_pkey_decrypt Encrypted data Raw data - 1728 kernel_pkey_sign Raw data Signature - 1729 kernel_pkey_verify Raw data - Signature 1730 1731 asym_eds_op() deals with encryption, decryption and signature creation as 1732 specified by params->op. Note that params->op is also set for 1733 asym_verify_signature(). 1734 1735 Encrypting and signature creation both take raw data in the input buffer 1736 and return the encrypted result in the output buffer. Padding may have 1737 been added if an encoding was set. In the case of signature creation, 1738 depending on the encoding, the padding created may need to indicate the 1739 digest algorithm - the name of which should be supplied in hash_algo. 1740 1741 Decryption takes encrypted data in the input buffer and returns the raw 1742 data in the output buffer. Padding will get checked and stripped off if 1743 an encoding was set. 1744 1745 Verification takes raw data in the input buffer and the signature in the 1746 second input buffer and checks that the one matches the other. Padding 1747 will be validated. Depending on the encoding, the digest algorithm used 1748 to generate the raw data may need to be indicated in hash_algo. 1749 1750 If successful, asym_eds_op() should return the number of bytes written 1751 into the output buffer. asym_verify_signature() should return 0. 1752 1753 A variety of errors may be returned, including EOPNOTSUPP if the operation 1754 is not supported; EKEYREJECTED if verification fails; ENOPKG if the 1755 required crypto isn't available. 1756 1757 1758 * ``int (*asym_query)(const struct kernel_pkey_params *params, 1759 struct kernel_pkey_query *info);`` 1760 1761 This method is optional. If provided it allows information about the 1762 public or asymmetric key held in the key to be determined. 1763 1764 The parameter block is as for asym_eds_op() and co. but in_len and out_len 1765 are unused. The encoding and hash_algo fields should be used to reduce 1766 the returned buffer/data sizes as appropriate. 1767 1768 If successful, the following information is filled in:: 1769 1770 struct kernel_pkey_query { 1771 __u32 supported_ops; 1772 __u32 key_size; 1773 __u16 max_data_size; 1774 __u16 max_sig_size; 1775 __u16 max_enc_size; 1776 __u16 max_dec_size; 1777 }; 1778 1779 The supported_ops field will contain a bitmask indicating what operations 1780 are supported by the key, including encryption of a blob, decryption of a 1781 blob, signing a blob and verifying the signature on a blob. The following 1782 constants are defined for this:: 1783 1784 KEYCTL_SUPPORTS_{ENCRYPT,DECRYPT,SIGN,VERIFY} 1785 1786 The key_size field is the size of the key in bits. max_data_size and 1787 max_sig_size are the maximum raw data and signature sizes for creation and 1788 verification of a signature; max_enc_size and max_dec_size are the maximum 1789 raw data and signature sizes for encryption and decryption. The 1790 max_*_size fields are measured in bytes. 1791 1792 If successful, 0 will be returned. If the key doesn't support this, 1793 EOPNOTSUPP will be returned. 1794 1795 1796Request-Key Callback Service 1797============================ 1798 1799To create a new key, the kernel will attempt to execute the following command 1800line:: 1801 1802 /sbin/request-key create <key> <uid> <gid> \ 1803 <threadring> <processring> <sessionring> <callout_info> 1804 1805<key> is the key being constructed, and the three keyrings are the process 1806keyrings from the process that caused the search to be issued. These are 1807included for two reasons: 1808 1809 1 There may be an authentication token in one of the keyrings that is 1810 required to obtain the key, eg: a Kerberos Ticket-Granting Ticket. 1811 1812 2 The new key should probably be cached in one of these rings. 1813 1814This program should set it UID and GID to those specified before attempting to 1815access any more keys. It may then look around for a user specific process to 1816hand the request off to (perhaps a path held in placed in another key by, for 1817example, the KDE desktop manager). 1818 1819The program (or whatever it calls) should finish construction of the key by 1820calling KEYCTL_INSTANTIATE or KEYCTL_INSTANTIATE_IOV, which also permits it to 1821cache the key in one of the keyrings (probably the session ring) before 1822returning. Alternatively, the key can be marked as negative with KEYCTL_NEGATE 1823or KEYCTL_REJECT; this also permits the key to be cached in one of the 1824keyrings. 1825 1826If it returns with the key remaining in the unconstructed state, the key will 1827be marked as being negative, it will be added to the session keyring, and an 1828error will be returned to the key requestor. 1829 1830Supplementary information may be provided from whoever or whatever invoked this 1831service. This will be passed as the <callout_info> parameter. If no such 1832information was made available, then "-" will be passed as this parameter 1833instead. 1834 1835 1836Similarly, the kernel may attempt to update an expired or a soon to expire key 1837by executing:: 1838 1839 /sbin/request-key update <key> <uid> <gid> \ 1840 <threadring> <processring> <sessionring> 1841 1842In this case, the program isn't required to actually attach the key to a ring; 1843the rings are provided for reference. 1844 1845 1846Garbage Collection 1847================== 1848 1849Dead keys (for which the type has been removed) will be automatically unlinked 1850from those keyrings that point to them and deleted as soon as possible by a 1851background garbage collector. 1852 1853Similarly, revoked and expired keys will be garbage collected, but only after a 1854certain amount of time has passed. This time is set as a number of seconds in:: 1855 1856 /proc/sys/kernel/keys/gc_delay 1857