Zhanna Tsitkov [Mon, 23 Dec 2013 14:23:54 +0000 (09:23 -0500)]
Make KDC "status" statements more homogeneous
Generally we want KDC status strings to be concise, informative and follow
some common rules:
- All letters in the status string should be capitalized;
- the words in the status phrase are separated by underscore;
- abbreviations should be avoided. Some acceptable "standard" acronyms
are AS_REQ, TGS_REP etc.
- since in almost all cases KDC status is set on error, no need
to state this fact as part of the status string;
- KDC status string should be an imperative phrase.
For example, "DECRYPT_SERVER_KEY".
This commit is to modify some KDC status messages to follow this format.
Even though KDC status messages are not standardized, it is possible that some
administrators use them in the Kerberos log file processing. Hence, the vast
majority of them are left unchanged pending further investigation (mostly,
feedback from the administrators).
Nate Rosenblum [Mon, 23 Dec 2013 21:21:44 +0000 (13:21 -0800)]
Support referrals from Windows Server 2003
Although RFC 6806 Section 7 requires servers to indicate a client
referral in a WRONG_REALM message, Microsoft Windows Server 2003
returns this information in a message with error code
PRINCIPAL_UNKNOWN. Failure to follow the referral in these messages
prevents referral chasing in Windows Server 2003 forests. Detect
referral messages of this type by checking for a non-empty
client.realm field in the response, and activate the referral logic in
these cases.
[tlyu@mit.edu: style, comments, and commit message]
Greg Hudson [Wed, 12 Feb 2014 03:16:56 +0000 (22:16 -0500)]
Don't use system dictionary files for DB2 tests
The system dictionary may contain entries with punctuation, which can
confuse the shell. It's more predictable to always use the word list
from the source tree.
Greg Hudson [Wed, 12 Feb 2014 01:00:51 +0000 (20:00 -0500)]
Fix SAM-2 preauth when password argument is used
sam2_process accesses gak_data to get the password, so that it can do
string-to-key with the etype in the SAM-2 challenge. When #7642
changed gic_pwd.c to use struct gak_password instead of krb5_data,
sam2_process wasn't altered to match. We don't see a problem when the
password is read through the prompter (as with kinit), because the
password winds up in the storage field at the beginning of the
gak_password structure. But when a password is supplied as a
parameter (as with ksu), the storage field is empty and we get the
wrong answer from sam2_process.
Greg Hudson [Mon, 10 Feb 2014 20:26:34 +0000 (15:26 -0500)]
Correctly get default realm in kdb5.c
Call krb5_get_default_realm instead of directly accessing
context->default_realm, to remove the requirement that
krb5_get_default_realm or krb5_set_default_realm be used before
krb5_db_open.
Some system configurations expect Unix-domain sockets to live under
/run or /var/run, and not other parts of /var where persistent
application state lives. Define a new directory KDC_RUN_DIR using
$runstatedir (new in autoconf 2.70, so fall back to $localstatedir/run
if it's not set) and use that for the default socket path.
Greg Hudson [Tue, 4 Feb 2014 01:59:54 +0000 (20:59 -0500)]
Test SPNEGO acceptor response to MS krb5 mech OID
In t_spnego.c, add code to make a SPNEGO request with the erroneous
Microsoft OID value and examine the response to make sure that it uses
the same OID value as the request did. The token and tmp variables
were unused, so rename them to itok and atok for the purpose of the
new test code.
Greg Hudson [Tue, 4 Feb 2014 02:11:34 +0000 (21:11 -0500)]
Properly reflect MS krb5 mech in SPNEGO acceptor
r25590 changed negotiate_mech() to return an alias into the acceptor's
mech set, with the unfortunate side effect of transforming the
erroneous Microsoft krb5 mech OID into the correct krb5 mech OID,
meaning that we answer with a different OID than the requested one.
Return an alias into the initiator's mech set instead, and store that
in mech_set field the SPNEGO context. The acceptor code only uses
mech_set to hold the allocated storage pointed into by internal_mech,
so this change is safe.
Greg Hudson [Thu, 30 Jan 2014 17:37:46 +0000 (12:37 -0500)]
Don't get KDB lock in ulog_get_entries
ulog_get_entries does not access the KDB, only the ulog, so it does
not need a KDB lock; its read lock on the ulog is sufficient to
prevent logged updates from happening while it is running. There is
no reason to serialize against unlogged KDB updates such as those
performed by the KDC, as those do not affect the ulog.
Greg Hudson [Thu, 23 Jan 2014 15:51:21 +0000 (10:51 -0500)]
In kdb5_util load, init ulog closer to promote
Always map the ulog with FKCOMMAND, not FKLOAD which reinitializes the
ulog. Don't reinitialize the ulog until just before calling
krb5_db_promote(). Get rid of FKLOAD since it isn't needed any more;
we can just call ulog_init_header() manually.
Greg Hudson [Sun, 26 Jan 2014 21:42:50 +0000 (16:42 -0500)]
Clean up ovsec_kadmd.c
Fix code formatting issues. Use static scope when possible, and put
all helper functions before main so that forward declarations aren't
needed. Use GSS_KRB5_NT_PRINCIPAL_NAME instead of constructing it
with gss_str_to_oid, and get rid of display_status as that was the
only use. Replace large verbose function comments with concise ones.
Factor out main loop setup and KDB keytab setup into helper functions.
Create a helper function fail_to_start for logging and exiting on
initialization failures, and use it instead of the large cleanup
blocks in main.
Some initialization error messages have changed; no other functional
changes.
Greg Hudson [Thu, 23 Jan 2014 23:15:24 +0000 (18:15 -0500)]
Modernize iprop code
* Don't use "extern" for kdb_log.h prototypes.
* Avoid passing structures by value.
* Avoid the need to cast the result of the INDEX macro, and use char *
instead of unsigned long for pointer arithmetic.
* Reorganize kdb_log.c so static helpers are at the top and don't use
the "ulog_" prefix.
* Get rid of ulog_finish_update_slave since it's more concise to
open-code it in ulog_replay.
* Get rid of ulog_delete_update. In krb5_db_delete_principal, just
call ulog_add_update with kdb_deleted set in upd.
* Modernize coding style of kproplog.c. Use k5memdup0 instead of
snprintf in print_str to convert a byte range to a C string.
Remove an unnecesary textdomain call; libkrb5 takes care of calling
bindtextdomain in the library initializer.
* Modernize coding style of kpropd.c and kprop.c.
Greg Hudson [Fri, 24 Jan 2014 04:49:34 +0000 (23:49 -0500)]
Remove kdb5_util load iprop safety net
Revert the safety net added #7370. As written it only applied to the
master KDC, where a normal load sometimes makes sense. Even on a
slave, there are edge cases where it can make sense to do a normal
load.
Greg Hudson [Sun, 26 Jan 2014 22:23:53 +0000 (17:23 -0500)]
Check for unstable ulog in ulog_get_entries
If a process terminates in the middle of a logged database update, it
will release its lock and leave the ulog in the KDB_UNSTABLE state.
kadmind should notice this when it calls ulog_get_entries, but right
now it only checks for the KDB_CORRUPT state (which is never set any
more) and does not recover. Check for any state other than KDB_STABLE
and recover by resetting the ulog and forcing full resyncs to slaves.
Greg Hudson [Thu, 30 Jan 2014 17:26:09 +0000 (12:26 -0500)]
Add test case for ulog serial number wrapping
Add a unit test program named t_ulog. Make it map a ulog file, set up
header state where kdb_last_sno is the last serial number, and then
add an empty update. Check that the result is a ulog header with one
update at serial number 1.
Greg Hudson [Tue, 21 Jan 2014 15:58:04 +0000 (10:58 -0500)]
Reinitialize ulog when wrapping serial number
When we wrap the serial number, reinitialize the ulog so that kdb_num
starts counting from 1, instead of leaving the ulog in a weird state
where kdb_num == ulogentries but kdb_last_sno == 1 and the serial
numbers in the circular array aren't necessarily contiguous.
Greg Hudson [Sat, 18 Jan 2014 18:03:32 +0000 (13:03 -0500)]
Fix gss_pseudo_random leak on zero length output
Nobody is likely to ever ask for zero bytes of output from
gss_pseudo_random, but if they do, just return an empty buffer without
allocating. Otherwise we leak memory because gss_release_buffer
doesn't do anything to buffers with length 0.
Greg Hudson [Thu, 16 Jan 2014 16:32:10 +0000 (11:32 -0500)]
Avoid assertion failure in error_message
r17942 added a call to get_thread_buffer in the first part of
error_message, prior to the call to com_err_initialize. This can
cause an assertion failure from k5_getspecific if error_message is
called on a system error before any other com_err functions are
called. Move the initialization call earlier to prevent this.
Greg Hudson [Wed, 15 Jan 2014 17:51:42 +0000 (12:51 -0500)]
Clean up GSS krb5 acquire_accept_cred
Use a cleanup handler instead of releasing kt in multiple error
clauses. Wrap a long line and fix a comment with a missing word.
Rewrap the function arguments to use fewer lines.
Tom Yu [Tue, 14 Jan 2014 20:43:35 +0000 (15:43 -0500)]
Remove mentions of krb5-send-pr
Start the process of deprecating krb5-send-pr. In practice, it causes
frustration for some users, and free-form email is good enough for
most bug reports.
Continue to install krb5-send-pr for now, but plan to remove it from
the tree in the future, probably replaced by a script that instructs
the user to send email manually.
Greg Hudson [Mon, 13 Jan 2014 17:02:09 +0000 (12:02 -0500)]
Don't produce context deletion token in krb5 mech
RFCs 2743 and 4121 recommend that implementations produce empty tokens
from gss_delete_sec_context, and trying to produce one can cause
gss_delete_sec_context to fail on a partially established context.
Patch from Tomas Kuthan.
Greg Hudson [Fri, 10 Jan 2014 16:54:13 +0000 (11:54 -0500)]
Restrict AES-NI support to ELF platforms for now
Since we explicitly specify the ELF object format when building
iaesx86.s or iaesx64.s, we need to restrict it to operating systems we
know to be ELF platforms. Otherwise we can break the build on OS X,
which uses the Mach-O object format.
Greg Hudson [Thu, 9 Jan 2014 05:18:44 +0000 (00:18 -0500)]
Work around Linux session keyring write behavior
If the session keyring matches the user session keyring, write
explicitly to the user session keyring. Otherwise the kernel might
create a new session keyring for the process, making the resulting
cache collection invisible to other processes.
Tom Yu [Tue, 7 Jan 2014 21:37:46 +0000 (16:37 -0500)]
Refactor krb5_string_to_keysalts()
Use various standard C library functions rather than rolling our own.
Previous code spent many lines reimplementing realloc(), strpbrk(),
strtok_r(), etc.
Make a separate string_to_keysalt() parser for an individual keysalt
pair, which for now is private and destructive.
Tom Yu [Mon, 6 Jan 2014 22:17:02 +0000 (17:17 -0500)]
Make salt defaulting work for keysalts
Make krb5_string_to_keysalts() default to only ":" as a key:salt
separator character. Change most of its callers to pass NULL so they
get the default separators.
Tom Yu [Mon, 6 Jan 2014 21:32:50 +0000 (16:32 -0500)]
Default to normal salt instead of "ignore"
krb5_string_to_keysalts() treats an empty salt field as -1 ("ignore"),
rather than as the normal salttype. Default to normal instead, so
that omitting a salttype works as expected.
Greg Hudson [Fri, 3 Jan 2014 18:50:48 +0000 (13:50 -0500)]
Mark AESNI files as not needing executable stacks
Some Linux systems now come with facilities to mark the stack as
non-executable, making it more difficult to exploit buffer overrun
bugs. For this to work, object files built from assembly need a
section added to note whether they require an executable stack.
Patch from Dhiru Kholia with comments added. More information at:
https://bugzilla.redhat.com/show_bug.cgi?id=1045699
https://wiki.gentoo.org/wiki/Hardened/GNU_stack_quickstart
Tom Yu [Wed, 1 Jan 2014 00:43:28 +0000 (19:43 -0500)]
Test bogus KDC-REQs
Send encodings that are invalid KDC-REQs, but pass krb5_is_as_req()
and krb5_is_tgs_req(), to make sure that the KDC recovers correctly
from failures in decode_krb5_as_req() and decode_krb5_tgs_req(). Also
send an encoding that isn't a valid KDC-REQ.
rbasch [Mon, 16 Dec 2013 15:54:41 +0000 (10:54 -0500)]
Log service princ in KDC more reliably
Under some error conditions, the KDC would log "<unknown server>" for
the service principal because service principal information is not yet
available to the logging functions. Set the appropriate variables
earlier.
do_as_req.c: After unparsing the client, immediately unparse the
server before searching for the client principal in the KDB.
do_tgs_req.c: Save a pointer to the client-requested service
principal, to make sure it gets logged if an error happens before
search_sprinc() successfully completes.
[tlyu@mit.edu: commit message; fix TGS to catch more error cases]
Greg Hudson [Thu, 19 Dec 2013 18:33:33 +0000 (13:33 -0500)]
Allow realm in kadm5_init service names
Previously, if you passed a service name with a realm part to a
kadm5_init function, you would get a KRB5_PARSE_MALFORMED error
because the code would internally append its own '@realm' suffix
before parsing the name. Fix this as follows:
Change gic_iter so instead of producing a full service name, it
produces a krb5_principal which is taken from the cred it acquires.
Pass the client and full service name around as principals, rather
than strings, and use the gss_nt_krb5_principal name type to import
them in setup_gss(). Don't append a realm to the input service name;
instead, pass the input service name directly to the gic functions
(which do not need a realm in the service name and will ignore the
realm if one is present). For the INIT_CREDS case, parse the input
service name with KRB5_PRINCIPAL_PARSE_IGNORE_REALM and then set the
realm.
Greg Hudson [Thu, 19 Dec 2013 17:22:47 +0000 (12:22 -0500)]
Simplify libkadm5 client realm initialization
The "realm" variable in init_any is used only to fill in the realm of
the service principal in gic_iter(). The service principal realm
should always be the realm we looked up config parameters for, so we
can supply that realm to get_init_creds() unconditionally and
eliminate the case where we use the client principal realm.
Also get rid of an outdated comment and an #if 0 block we will never
need again, and use SNPRINTF_OVERFLOW to check the snprintf result.
Greg Hudson [Fri, 20 Dec 2013 16:06:52 +0000 (11:06 -0500)]
Use an extended com_err hook in klist
Add an adapted version of extended_com_err_fn from kinit to klist and
use it. In do_ccache(), rely on the ccache type to set a reasonable
message if krb5_cc_set_flags() or krb5_cc_get_principal() fails due to
a nonexistent or unreadable ccache, and don't confuse the user with
the name of the ccache operation that failed.
Nalin Dahyabhai [Thu, 5 Dec 2013 18:54:09 +0000 (13:54 -0500)]
Set an error message when keyring get_princ fails
When attempting to use a keyring cache that doesn't exist, set an error
message when we fail to read a principal name, as we do when we return
the same error code when using a file ccache.
[ghudson: removed unnecessary check for d->name nullity.]
Greg Hudson [Fri, 20 Dec 2013 04:47:22 +0000 (23:47 -0500)]
Test for verto_set_flags in system libverto
libkrad relies on verto_set_flags, which was added to libverto in
release 0.2.4. Make sure the system libverto has this function before
choosing it over the built-in version.
Zhanna Tsitkov [Thu, 19 Dec 2013 18:08:56 +0000 (13:08 -0500)]
Move kprop error explanation into Troubleshooting
The plan is to make Troubleshooting section of the documentation a
one-stop-shop place for all error diagnostics, explanations and possible
solutions. The relocation of kprop error messages descriptions is part of
this consolidation effort.
Greg Hudson [Wed, 18 Dec 2013 18:08:25 +0000 (13:08 -0500)]
Add a test program for krb5_copy_context
This test program isn't completely proof against the kind of mistakes
we've made with krb5_copy_context in the past, but it at least
exercises krb5_copy_context and can detect some kinds of bugs.
Greg Hudson [Wed, 18 Dec 2013 20:03:03 +0000 (15:03 -0500)]
Fix krb5_copy_context
krb5_copy_context has been broken since 1.8 (it broke in r22456)
because k5_copy_etypes crashes on null enctype lists. Subsequent
additions to the context structure were not reflected in
krb5_copy_context, creating double-free bugs. Make k5_copy_etypes
handle null input and account for all new fields in krb5_copy_context.
Reported by Arran Cudbard-Bell.
Simo Sorce [Tue, 17 Dec 2013 21:15:14 +0000 (16:15 -0500)]
Let SPNEGO display mechanism errors
To avoid potential recursion we use a thread local variable that tells
us whether the ancestor was called via spnego_gss_display_name(). If
we detect recursion, we assume that we returned a com_err code like
ENOMEM and call error_message(); in the worst case that will result in
an "Unknown error" message.
[ghudson@mit.edu: Edited comments and commit message; removed an
unneeded line of code.]
Greg Hudson [Tue, 17 Dec 2013 21:56:41 +0000 (16:56 -0500)]
Clarify klist -s documentation
The documentation for klist -s erroneously suggests that it doesn't
affect the exit status behavior and that it merely checks for the
existence of the ccache (only mentioning the expired ticket check at
the end). Make it clearer and simpler, but avoid going into a lot of
detail about the nature of the expiration check.
Greg Hudson [Mon, 16 Dec 2013 22:09:00 +0000 (17:09 -0500)]
Don't require krb5.conf without KRB5_DNS_LOOKUP
For a long time we have allowed krb5 contexts to be initialized in the
absence of krb5.conf--but only if KRB5_DNS_LOOKUP is defined,
presumably on the theory that no KDCs could be contacted without
either DNS support or profile configuration. But locate plugins could
provide the ability to find KDCs, and some libkrb5 operations (such as
IAKERB initiation) could succeed without needing to locate KDCs.
Also get rid of the profile_in_memory context flag, since we don't use
it any more.
Greg Hudson [Mon, 16 Dec 2013 20:37:56 +0000 (15:37 -0500)]
Fix GSS krb5 acceptor acquire_cred error handling
When acquiring acceptor creds with a specified name, if we fail to
open a replay cache, we leak the keytab handle. If there is no
specified name and we discover that there is no content in the keytab,
we leak the keytab handle and return the wrong major code. Memory
leak reported by Andrea Campi.
Simo Sorce [Fri, 13 Dec 2013 17:00:41 +0000 (12:00 -0500)]
Fix memory leak in SPNEGO initiator
If we eliminate a mechanism from the initiator list because
gss_init_sec_context fails, free the memory for that mech OID before
removing it from the list.
Greg Hudson [Mon, 16 Dec 2013 16:35:42 +0000 (11:35 -0500)]
Remove unneeded check in SPNEGO initiator
In init_ctx_cont, if the response token contains no fields, we set a
return value but don't actually quit out of the function. We do not
need this check (we will fail later on if a piece of required
information isn't present), so just remove it. Reported by
simo@redhat.com.
Greg Hudson [Tue, 10 Dec 2013 17:04:18 +0000 (12:04 -0500)]
Fix SPNEGO one-hop interop against old IIS
IIS 6.0 and similar return a zero length reponse buffer in the last
SPNEGO packet when context initiation is performed without mutual
authentication. In this case the underlying Kerberos mechanism has
already completed successfully on the first invocation, and SPNEGO
does not expect a mech response token in the answer. If we get an
empty mech response token when the mech is complete during
negotiation, ignore it.
[ghudson@mit.edu: small code style and commit message changes]
Greg Hudson [Sun, 8 Dec 2013 23:05:26 +0000 (18:05 -0500)]
Allow ":port" suffixes in sn2princ hostnames
MSSQLSvc principal names can contain a ":port" or ":instance" trailer
on the hostname part. If we see that in the hostname argument of
krb5_sname_to_principal(), remove it before canonicalizing the
hostname and put it back on afterwards.
Greg Hudson [Fri, 6 Dec 2013 01:32:05 +0000 (20:32 -0500)]
Fix S4U2Self against non-FAST KDCs
When we added FAST TGS support in 1.11, we broke S4U2Self against KDCs
which don't support FAST, because the S4U2Self padata is only present
within the FAST request. For now, duplicate that padata in the outer
request so that both FAST and non-FAST KDCs can see it.
Greg Hudson [Tue, 3 Dec 2013 07:05:46 +0000 (02:05 -0500)]
Edit README.asn1
Add another blank line before section headers. Avoid contractions.
Change some whiches to thats where it seems appropriate. Fix some
missing or extra words.
Greg Hudson [Mon, 25 Nov 2013 16:46:47 +0000 (11:46 -0500)]
Correctly log IPv6 addresses in kadmind
Define client_addr() in server_stubs.c and use it consistently in that
file and ipropd_svc.c to get the client address from a transport
handle. In it, call getpeername() on the client socket and use
inet_ntop() on the result, instead of using inet_ntoa() on the IPv4
socket address. Provide a log_badauth2 callback to GSSRPC, so that we
get a transport handle instead of an IPv4 socket address, and use
client_addr() within it instead of inet_ntoa().
Greg Hudson [Mon, 25 Nov 2013 16:33:35 +0000 (11:33 -0500)]
Add new versions of log_badauth gssrpc callbacks
libgssrpc supports two callbacks for gss_accept_sec_context failures
on servers (one for AUTH_GSS and one for AUTH_GSSAPI), which are
IPv4-specific. Provide an alternate version which supplies the
transport handle instead of the address, so that we can get the
address via the file descriptor for TCP connections.
Greg Hudson [Thu, 21 Nov 2013 22:30:54 +0000 (17:30 -0500)]
Improve default ccache name API documentation
Document the lifetime and caching behavior of the
krb5_cc_default_name() return value. Document that
krb5_cc_set_default_name() may be called with NULL to purge the cached
value. Correct a typo in the krb5_cc_default() summary and explicitly
reference krb5_cc_default_name().
Greg Hudson [Thu, 21 Nov 2013 21:18:27 +0000 (16:18 -0500)]
Add another kadmin ACL test for backreferences
Add a test using backreferences which don't correspond directly to
principal components, to verify that *N refers to the Nth wildcard and
not the Nth component.
Greg Hudson [Mon, 18 Nov 2013 23:59:17 +0000 (18:59 -0500)]
Clarify lockout replication issues in docs
In the "KDC replication and account lockout" section of lockout.rst,
specifically call out kprop and incremental propagation as the
mechanisms which do not replicate account lockout state, and add a
note that KDCs using LDAP may not be affected by that section's
concerns.
Greg Hudson [Sun, 17 Nov 2013 17:37:09 +0000 (12:37 -0500)]
Remove dangling --with-kdc-kdb-update references
This configure option hasn't done anything since 1.8, so don't mention
it in configure --help or the documentation. The disable_last_success
and disable_lockout DB options are now used to turn it off.
Greg Hudson [Sat, 16 Nov 2013 04:38:15 +0000 (23:38 -0500)]
Remove a warning in AES string-to-key
On 32-bit platforms, the code to translate an iteration count of 0 to
2^32 can trigger a compiler warning. Since we will basically never
accept an iteration count that high (right now we reject anything
above 2^24), just reject it out of hand.
Simo Sorce [Thu, 14 Nov 2013 22:23:59 +0000 (17:23 -0500)]
Add support to store time offsets in cc_keyring
The code follows the same model used for the memory ccache type. Time
offsets are stored in each credential cache in a special key just like
the principal name. Legacy session caches do not store timestamps as
legacy code would fail when iterating over the new offset key.
[ghudson@mit.edu: minor formatting changes; note legacy session
exception in commit message]