Documentation/keys-request-key.txt

   1                               ===================
   2                               KEY REQUEST SERVICE
   3                               ===================
   4
   5 The key request service is part of the key retention service (refer to
   6 Documentation/keys.txt).  This document explains more fully how the requesting
   7 algorithm works.
   8
   9 The process starts by either the kernel requesting a service by calling
  10 request_key*():
  11
  12         struct key *request_key(const struct key_type *type,
  13                                 const char *description,
  14                                 const char *callout_string);
  15
  16 or:
  17
  18         struct key *request_key_with_auxdata(const struct key_type *type,
  19                                              const char *description,
  20                                              const char *callout_string,
  21                                              void *aux);
  22
  23 or:
  24
  25         struct key *request_key_async(const struct key_type *type,
  26                                       const char *description,
  27                                       const char *callout_string);
  28
  29 or:
  30
  31         struct key *request_key_async_with_auxdata(const struct key_type *type,
  32                                                    const char *description,
  33                                                    const char *callout_string,
  34                                                    void *aux);
  35
  36 Or by userspace invoking the request_key system call:
  37
  38         key_serial_t request_key(const char *type,
  39                                  const char *description,
  40                                  const char *callout_info,
  41                                  key_serial_t dest_keyring);
  42
  43 The main difference between the access points is that the in-kernel interface
  44 does not need to link the key to a keyring to prevent it from being immediately
  45 destroyed.  The kernel interface returns a pointer directly to the key, and
  46 it's up to the caller to destroy the key.
  47
  48 The request_key*_with_auxdata() calls are like the in-kernel request_key*()
  49 calls, except that they permit auxiliary data to be passed to the upcaller (the
  50 default is NULL).  This is only useful for those key types that define their
  51 own upcall mechanism rather than using /sbin/request-key.
  52
  53 The two async in-kernel calls may return keys that are still in the process of
  54 being constructed.  The two non-async ones will wait for construction to
  55 complete first.
  56
  57 The userspace interface links the key to a keyring associated with the process
  58 to prevent the key from going away, and returns the serial number of the key to
  59 the caller.
  60
  61
  62 The following example assumes that the key types involved don't define their
  63 own upcall mechanisms.  If they do, then those should be substituted for the
  64 forking and execution of /sbin/request-key.
  65
  66
  67 ===========
  68 THE PROCESS
  69 ===========
  70
  71 A request proceeds in the following manner:
  72
  73  (1) Process A calls request_key() [the userspace syscall calls the kernel
  74      interface].
  75
  76  (2) request_key() searches the process's subscribed keyrings to see if there's
  77      a suitable key there.  If there is, it returns the key.  If there isn't,
  78      and callout_info is not set, an error is returned.  Otherwise the process
  79      proceeds to the next step.
  80
  81  (3) request_key() sees that A doesn't have the desired key yet, so it creates
  82      two things:
  83
  84      (a) An uninstantiated key U of requested type and description.
  85
  86      (b) An authorisation key V that refers to key U and notes that process A
  87          is the context in which key U should be instantiated and secured, and
  88          from which associated key requests may be satisfied.
  89
  90  (4) request_key() then forks and executes /sbin/request-key with a new session
  91      keyring that contains a link to auth key V.
  92
  93  (5) /sbin/request-key assumes the authority associated with key U.
  94
  95  (6) /sbin/request-key execs an appropriate program to perform the actual
  96      instantiation.
  97
  98  (7) The program may want to access another key from A's context (say a
  99      Kerberos TGT key).  It just requests the appropriate key, and the keyring
 100      search notes that the session keyring has auth key V in its bottom level.
 101
 102      This will permit it to then search the keyrings of process A with the
 103      UID, GID, groups and security info of process A as if it was process A,
 104      and come up with key W.
 105
 106  (8) The program then does what it must to get the data with which to
 107      instantiate key U, using key W as a reference (perhaps it contacts a
 108      Kerberos server using the TGT) and then instantiates key U.
 109
 110  (9) Upon instantiating key U, auth key V is automatically revoked so that it
 111      may not be used again.
 112
 113 (10) The program then exits 0 and request_key() deletes key V and returns key
 114      U to the caller.
 115
 116 This also extends further.  If key W (step 7 above) didn't exist, key W would
 117 be created uninstantiated, another auth key (X) would be created (as per step
 118 3) and another copy of /sbin/request-key spawned (as per step 4); but the
 119 context specified by auth key X will still be process A, as it was in auth key
 120 V.
 121
 122 This is because process A's keyrings can't simply be attached to
 123 /sbin/request-key at the appropriate places because (a) execve will discard two
 124 of them, and (b) it requires the same UID/GID/Groups all the way through.
 125
 126
 127 ======================
 128 NEGATIVE INSTANTIATION
 129 ======================
 130
 131 Rather than instantiating a key, it is possible for the possessor of an
 132 authorisation key to negatively instantiate a key that's under construction.
 133 This is a short duration placeholder that causes any attempt at re-requesting
 134 the key whilst it exists to fail with error ENOKEY.
 135
 136 This is provided to prevent excessive repeated spawning of /sbin/request-key
 137 processes for a key that will never be obtainable.
 138
 139 Should the /sbin/request-key process exit anything other than 0 or die on a
 140 signal, the key under construction will be automatically negatively
 141 instantiated for a short amount of time.
 142
 143
 144 ====================
 145 THE SEARCH ALGORITHM
 146 ====================
 147
 148 A search of any particular keyring proceeds in the following fashion:
 149
 150  (1) When the key management code searches for a key (keyring_search_aux) it
 151      firstly calls key_permission(SEARCH) on the keyring it's starting with,
 152      if this denies permission, it doesn't search further.
 153
 154  (2) It considers all the non-keyring keys within that keyring and, if any key
 155      matches the criteria specified, calls key_permission(SEARCH) on it to see
 156      if the key is allowed to be found.  If it is, that key is returned; if
 157      not, the search continues, and the error code is retained if of higher
 158      priority than the one currently set.
 159
 160  (3) It then considers all the keyring-type keys in the keyring it's currently
 161      searching.  It calls key_permission(SEARCH) on each keyring, and if this
 162      grants permission, it recurses, executing steps (2) and (3) on that
 163      keyring.
 164
 165 The process stops immediately a valid key is found with permission granted to
 166 use it.  Any error from a previous match attempt is discarded and the key is
 167 returned.
 168
 169 When search_process_keyrings() is invoked, it performs the following searches
 170 until one succeeds:
 171
 172  (1) If extant, the process's thread keyring is searched.
 173
 174  (2) If extant, the process's process keyring is searched.
 175
 176  (3) The process's session keyring is searched.
 177
 178  (4) If the process has assumed the authority associated with a request_key()
 179      authorisation key then:
 180
 181      (a) If extant, the calling process's thread keyring is searched.
 182
 183      (b) If extant, the calling process's process keyring is searched.
 184
 185      (c) The calling process's session keyring is searched.
 186
 187 The moment one succeeds, all pending errors are discarded and the found key is
 188 returned.
 189
 190 Only if all these fail does the whole thing fail with the highest priority
 191 error.  Note that several errors may have come from LSM.
 192
 193 The error priority is:
 194
 195         EKEYREVOKED > EKEYEXPIRED > ENOKEY
 196
 197 EACCES/EPERM are only returned on a direct search of a specific keyring where
 198 the basal keyring does not grant Search permission.