This is the second of two commits in a series that must be picked together.
This series of two commits moves parts of lib/krb5/ infrastructure
functionality to lib/base/, leaving behind wrappers.
Some parts of libkrb5 are entirely generic or easily made so, and could
be useful in various parts of Heimdal that are not specific to the krb5
API, such as:
- lib/gssapi/ (especially since the integration of NegoEx)
- lib/hx509/
- bx509d (which should really move out of kdc/)
For the above we need to move these bits of lib/krb5/:
- lib/krb5/config_file.c (all of it, leaving forwardings behind)
- lib/krb5/config_reg.c (all of it)
- lib/krb5/plugin.c (all of it, leaving forwardings behind)
- lib/krb5/log.c (all of it, ditto)
- lib/krb5/heim_err.et (all of it)
And because of those two, these too must also move:
- lib/krb5/expand_path.c (all of it, leaving forwardings behind)
- lib/krb5/warn.c (just the warning functions, ditto)
The changes to the moved files are mostly quite straightforward and are
best reviewed with --word-diff=color.
We're also creating a heim_context and a heim API to go with it. But
it's as thin as possible, with as little state as necessary to enable
this move. Functions for dealing with error messages use callbacks.
Moving plugin.c does have one knock-on effect on all users of the old
krb5 plugin API (which remains), which is that a global search and
replace of struct krb5_plugin_data to struct heim_plugin_data was
needed, though the layout and size of that structure doesn't change, so
the ABI doesn't either.
As well, we now build lib/vers/ and lib/com_err/ before lib/base/ so as
to be able to move lib/krb5/heim_err.et to lib/base/ so that we can make
use of HEIM_ERR_* in lib/base/, specifically in the files that moved.
Once this is all done we'll be able to use config files and plugins in
lib/hx509/, we'll be able to move bx509d out of kdc/, and so on.
Most if not all of the new functions in lib/base/ are Heimdal-private,
thus calling conventions for them are not declared.
Status:
- builds and passes CIs (Travis, Appveyor)
- ran make check-valgrind and no new leaks or other memory errors
- ready for review
HOW TO REVIEW:
$ # Review file moves:
$ git log --stat -n1 HEAD^
$
$ # Review changes to moved files using --word-diff=color
$ git log -p -b -w --word-diff=color HEAD^..HEAD \
lib/base/config_file.c \
lib/base/config_reg.c \
lib/base/expand_path.c \
lib/base/warn.c \
lib/krb5/config_file.c \
lib/krb5/config_reg.c \
lib/krb5/expand_path.c \
lib/krb5/warn.c
$
$ # Review the whole thing, possibly adding -b and/or -w, and
$ # maybe --word-diff=color:
$ git log -p origin/master..HEAD
$ git log -p -b -w origin/master..HEAD
$ git log -p -b -w --word-diff=color origin/master..HEAD
TBD (future commits):
- make lib/gssapi use the new heimbase functions
- move kx509/bx509d common code to lib/hx509/ or other approp. location
- move bx509d out of kdc/
Refactor plugin framework to use a single list of loaded plugins; add a new
plugin API where DSOs export a load function that can declare dependencies and
export multiple plugins; refactor kadm5 hook API to use krb5 plugin framework.
More information in krb5-plugin(7).
This change adds plugin support to the kadmin libraries for performing
actions before and after a password change is committed to the KDC database
and after a change is made to the attributes of a principal (specifically,
a change to DISALLOW_ALL_TIX).
This change adds a hook_libraries configuration option to the [kadmin]
section of krb5.conf (or kdc.conf if you use that file) that must be set
to load the module. That configuration option is in the form:
[kadmin]
hook_libraries = /usr/local/lib/krb5/plugins/kadm5_hook/krb5_sync.so
where the value is the full path to the plugin that you want to load. If
this option is not present, kadmind will not load a plugin and the changes
from the patch will be inactive. If this option is given and the plugin
cannot be loaded, kadmind startup will abort with a (hopefully useful)
error message in syslog.
Any plugin used with this patch must expose a public function named
kadm5_hook_init of type kadm5_hook_init_t that returns a kadm5_hook structure.
See sample_hook.c for an example of this initialization function.
typedef struct kadm5_hook {
const char *name;
uint32_t version;
const char *vendor;
void (KRB5_CALLCONV *fini)(krb5_context, void *data);
krb5_error_code (KRB5_CALLCONV *chpass)(krb5_context context,
void *data,
enum kadm5_hook_stage stage,
krb5_error_code code,
krb5_const_principal princ,
uint32_t flags,
size_t n_ks_tuple,
krb5_key_salt_tuple *ks_tuple,
const char *password,
char **error_msg);
...
};
where enum kadm5_hook_stage is:
enum kadm5_hook_stage {
KADM5_HOOK_STAGE_PRECOMMIT,
KADM5_HOOK_STAGE_POSTCOMMIT
};
init creates a hook context that is passed into all subsequent calls.
chpass is called for password changes, create is called for principal
creation (with the newly-created principal in the kadm5_principal_ent_t
argument), and modify is called when a principal is modified. The purpose of
the remaining functions should be self-explanatory.
returning 0 on success and a Kerberos error code on failure, setting the
Kerberos error message in the provided context. The error code passed in is
valid for post-commit hooks and contains the result of the update operation.
This change is submitted under the following license
Copyright 2012, 2013
The Board of Trustees of the Leland Stanford Junior University
Portions Copyright 2018 AuriStor Inc.
Copying and distribution of this file, with or without modification, are
permitted in any medium without royalty provided the copyright notice and
this notice are preserved. This file is offered as-is, without any
warranty.
When new keys are added (typically via kadm5_setkey_principal_3),
truncate the key history to remove old keys, that is keys older than
the newest key which was in effect prior longer ago than the principal's
maximum ticket lifetime. This feature is controlled via the "[kadmin]"
section's "prune-key-history" boolean parameter, which defaults to false.
Currently this happens only when kadm5_setkey_principal_3()
is called directly on the server, the client API simulates
kadm5_setkey_principal_3() via a get, update, modify sequence that does
not prune the key history. The plan is to add a new kadm5 protocol RPC
and convert clients to call that instead.
In setkey_principal_3 seal keys after entry key update
Also, for now, don't check the return value of kadm5_log_modify() in
the new kadm5_s_setkey_principal_3(). This has to be addressed more
globally.
Censor stale keys in kadm5_s_get_principal
We used to update the iprop log and HDB in different orders depending on
the kadm5 operation, which then led to various race conditions.
The iprop log now functions as a two-phase commit (with roll forward)
log for HDB changes. The log is auto-truncated, keeping the latest
entries that fit in a configurable maximum number of bytes (defaults to
50MB). See the log-max-size parameter description in krb5.conf(5).
The iprop log format and the protocol remain backwards-compatible with
earlier versions of Heimdal. This is NOT a flag-day; there is NO need
to update all the slaves at once with the master, though it is advisable
in general. Rolling upgrades and downgrades should work.
The sequence of updates is now (with HDB and log open and locked):
a) check that the HDB operation will succeed if attempted,
b) append to iprop log and fsync() it,
c) write to HDB (which should fsync()),
d) mark last log record committed (no fsync in this case).
Every kadm5 write operation recover transactions not yet confirmed as
committed, thus there can be at most one unconfirmed commit on a master
KDC.
Reads via kadm5_get_principal() also attempt to lock the log, and if
successful, recover unconfirmed transactions; readers must have write
access and must win any race to lock the iprop log.
The ipropd-master daemon also attempts to recover unconfirmed
transactions when idle.
The log now starts with a nop record whose payload records the offset of
the logical end of the log: the end of the last confirmed committed
transaction. This is kown as the "uber record". Its purpose is
two-fold: act as the confirmation of committed transactions, and provide
an O(1) method of finding the end of the log (i.e., without having to
traverse the entire log front to back).
Two-phase commit makes all kadm5 writes single-operation atomic
transactions (though some kadm5 operations, such as renames of
principals, and changes to principals' aliases, use multiple low-level
HDB write operations, but still all in one transaction). One can still
hold a lock on the HDB across many operations (e.g., by using the lock
command in a kadmin -l or calling kadm5_lock()) in order to push
multiple transactions in sequence, but this sequence will not be atomic
if the process or host crashes in the middle.
As before, HDB writes which do not go through the kadm5 API are excluded
from all of this, but there should be no such writes.
Lastly, the iprop-log(1) command is enhanced as follows:
- The dump, last-version, truncate, and replay sub-commands now have an
option to not lock the log. This is useful for inspecting a running
system's log file, especially on slave KDCs.
- The dump, last-version, truncate, and replay sub-commands now take an
optional iprop log file positional argument, so that they may be used
to inspect log files other than the running system's
configured/default log file.
Extensive code review and some re-writing for clarity by Viktor Dukhovni.
On Windows a file descriptor is an int value allocated by the
local module instance of the C Run Time Library. A socket handle is a
SOCKET value allocated by a Winsock Provider for the requested family and
protocol. These two values cannot be mixed and there is no mechanism for
converting between the two. The _get_osfhandle() and _open_osfhandle()
functions can work with a standard HANDLE (file, pipe, etc) but cannot be
used for a SOCKET.
The Heimdal krb5_storage_from_fd() routine counted on the osf conversion
functions working on SOCKET values. Since they do not any attempt to call
krb5_storage_from_fd() on a socket resulted in an assertion being thrown
by the C RTL.
Another problem is SOCKET value truncation when storing a 64-bit value
into a 32-bit int.
To address these problems a new krb5_storage_from_socket() routine is
introduced. This routine setups a krb5_storage that stores a socket value
as a rk_socket_t and provides a set of helper routines that always use
network ready functions.
The krb5_storage_from_fd() routines no longer use net_read() and
net_write() but provide helpers that follow their logic so that pipes can
be processed.
All call sites that allocate a socket now store the socket as rk_socket_t
and call krb5_storage_from_socket().
All locations that previously called the bare close() on a socket value
now call rk_closesocket().
Change-Id: I045f775b2a5dbf5cf803751409490bc27fffe597
We enable kadm5_chpass_principal_3() in the server side of the
library. The client kadm5 library calls will still return the
error KAMD5_KS_TUPLE_NO_SUPP.
Signed-off-by: Nicolas Williams <nico@cryptonector.com>
The libkadm5 functions hdb_open() and close around all HDB ops. This
meant the previous implementation of kadm5_lock() and unlock would
always result in a core dump. Now we hdb_open() for write in
kadm5_lock() and hdb_close() in kadm5_unlock(), with all kadm5_s_*()
functions now not opening nor closing the HDB when the server context
keep_open flag is set.
Also, there's now kadmin(8) lock and unlock commands. These are there
primarily as a way to test the kadm5_lock()/unlock() operations, but
MIT's kadmin.local also has lock/unlock commands, and these can be
useful for scripting (though they require much care).
