krb5.git
12 years agoMark Davies reported the password expiry message bug
Tom Yu [Thu, 8 Mar 2012 17:34:12 +0000 (17:34 +0000)]
Mark Davies reported the password expiry message bug

ticket: 7098
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25753 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoAdam Bernstein reported this bug and helped with analysis
Tom Yu [Thu, 8 Mar 2012 17:34:08 +0000 (17:34 +0000)]
Adam Bernstein reported this bug and helped with analysis

ticket: 7092
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25752 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25730 from trunk
Tom Yu [Wed, 7 Mar 2012 23:44:36 +0000 (23:44 +0000)]
Pull up r25730 from trunk

 ------------------------------------------------------------------------
 r25730 | ghudson | 2012-03-05 12:35:14 -0500 (Mon, 05 Mar 2012) | 15 lines

 ticket: 7098
 subject: Fix spurious password expiry warning
 target_version: 1.9.4
 tags: pullup

 r24241 (#6755) introduced a bug where if the KDC sends a LastReq entry
 containing an account expiry time, we send a prompter warning for
 password expiry even if there was no entry containing a password
 expiry time.  Typically, this results in the message "Warning: Your
 password will expire in less than one hour on Thu Jan  1 12:00:00
 1970".

 Fix this by explicitly checking for pw_exp == 0 in warn_pw_expiry()
 after we've gotten past the conditional for invoking the callback.

ticket: 7098
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25741 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25723 from trunk
Tom Yu [Wed, 7 Mar 2012 23:44:32 +0000 (23:44 +0000)]
Pull up r25723 from trunk

 ------------------------------------------------------------------------
 r25723 | ghudson | 2012-03-01 15:49:17 -0500 (Thu, 01 Mar 2012) | 16 lines

 ticket: 7096
 subject: Fix KDB iteration when callback does write calls
 target_version: 1.10.1
 tags: pullup

 kdb_db2's ctx_iterate makes an convenience alias to dbc->db in order
 to call more invoke call the DB's seq method.  This alias may become
 invalidated if the callback writes to the DB, since ctx_lock() may
 re-open the DB in order to acquire a write lock.  Fix the bug by
 getting rid of the convenience alias.

 Most KDB iteration operations in the code base do not write to the DB,
 but kdb5_util update_princ_encryption does.

 Bug discovered and diagnosed by will.fiveash@oracle.com.

ticket: 7096
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25740 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25725 from trunk, along with backport of r25703
Tom Yu [Wed, 7 Mar 2012 23:44:28 +0000 (23:44 +0000)]
Pull up r25725 from trunk, along with backport of r25703

 ------------------------------------------------------------------------
 r25725 | tlyu | 2012-03-02 17:24:38 -0500 (Fri, 02 Mar 2012) | 4 lines

 ticket: 7092

 Add test cases for Windows RODC kvno compatibility

 ------------------------------------------------------------------------
 r25703 | ghudson | 2012-02-21 13:57:44 -0500 (Tue, 21 Feb 2012) | 15 lines

 ticket: 7092
 subject: kvno ASN.1 encoding interop with Windows RODCs

 RFC 4120 defines the EncryptedData kvno field as an integer in the
 range of unsigned 32-bit numbers.  Windows encodes and decodes the
 field as a signed 32-bit integer.  Historically we do the same in our
 encoder in 1.6 and prior, and in our decoder through 1.10.  (Actually,
 our decoder through 1.10 decoded the value as a long and then cast the
 result to unsigned int, so it would accept positive values >= 2^31 on
 64-bit platforms but not on 32-bit platforms.)

 kvno values that large (or negative) are only likely to appear in the
 context of Windows read-only domain controllers.  So do what Windows
 does instead of what RFC 4120 says.

ticket: 7092
status: resolved
version_fixed: 1.10.1

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25739 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25716 from trunk
Tom Yu [Wed, 7 Mar 2012 23:44:20 +0000 (23:44 +0000)]
Pull up r25716 from trunk

 ------------------------------------------------------------------------
 r25716 | tlyu | 2012-02-27 13:31:50 -0500 (Mon, 27 Feb 2012) | 6 lines

 ticket: 7074
 target_version: 1.10.1
 tags: pullup

 Patch from Richard Basch to work around Solaris 8 lacking isblank().

ticket: 7074
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25738 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25704 from trunk
Tom Yu [Wed, 22 Feb 2012 04:11:56 +0000 (04:11 +0000)]
Pull up r25704 from trunk

 ------------------------------------------------------------------------
 r25704 | ghudson | 2012-02-21 14:14:47 -0500 (Tue, 21 Feb 2012) | 15 lines

 ticket: 7093
 subject: Access controls for string RPCs [CVE-2012-1012]
 target_version: 1.10.1
 tags: pullup

 In the kadmin protocol, make the access controls for
 get_strings/set_string mirror those of get_principal/modify_principal.
 Previously, anyone with global list privileges could get or modify
 string attributes on any principal.  The impact of this depends on how
 generous the kadmind acl is with list permission and whether string
 attributes are used in a deployment (nothing in the core code uses
 them yet).

 CVSSv2 vector: AV:N/AC:M/Au:S/C:P/I:P/A:N/E:H/RL:O/RC:C

ticket: 7093
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25709 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25700 from trunk
Tom Yu [Wed, 22 Feb 2012 04:11:51 +0000 (04:11 +0000)]
Pull up r25700 from trunk

 ------------------------------------------------------------------------
 r25700 | ghudson | 2012-02-14 13:26:50 -0500 (Tue, 14 Feb 2012) | 9 lines

 ticket: 7089
 subject: Initialize gss_get_name_attribute output buffers
 target_version: 1.10.1
 tags: pullup

 GSS functions which accept gss_buffer_t output arguments should
 initialize those buffers to empty on startup (see gss_wrap, for
 instance).  Do so for gss_get_name_attribute in the mechglue.

ticket: 7089
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25708 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25674 from trunk
Tom Yu [Wed, 22 Feb 2012 04:11:45 +0000 (04:11 +0000)]
Pull up r25674 from trunk

 ------------------------------------------------------------------------
 r25674 | ghudson | 2012-02-06 18:19:08 -0500 (Mon, 06 Feb 2012) | 18 lines

 ticket: 7087
 status: open
 target_version: 1.10.1
 tags: pullup

 Set display_value in krb5_gss_get_name_attribute

 A backwards conditional in r25358 caused krb5_gss_get_name_attribute
 not to set display_value on success.  Fix the sense of the
 conditional.

 We still don't quite correctly handle the cases where data_to_gss()
 fails, but those should be rare and the problem in those cases isn't
 severe, so it can be fixed separately.

 Also, value and display_value should probably be initialized to null
 buffers on failure, as is common with GSS interfaces.

ticket: 7087
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25707 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25679 from trunk
Tom Yu [Fri, 10 Feb 2012 21:19:13 +0000 (21:19 +0000)]
Pull up r25679 from trunk

 ------------------------------------------------------------------------
 r25679 | ghudson | 2012-02-07 10:45:59 -0500 (Tue, 07 Feb 2012) | 9 lines

 ticket: 7088
 subject: Fix uninitialized variable warning in trval.c
 target_version: 1.10.1
 tags: pullup

 gcc -O3 is agressive enough to detect that do_cons() doesn't always
 set the value of rlen2, so trval2() gets a maybe-uninitialized
 warning.  Squash it by initializing rlen2.

ticket: 7088
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25683 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25666 from trunk
Tom Yu [Fri, 10 Feb 2012 21:19:10 +0000 (21:19 +0000)]
Pull up r25666 from trunk

 ------------------------------------------------------------------------
 r25666 | ghudson | 2012-01-31 16:35:34 -0500 (Tue, 31 Jan 2012) | 12 lines

 ticket: 7084
 subject: Don't check mech in krb5_gss_inquire_cred_by_mech
 target_version: 1.10.1
 tags: pullup

 krb5_gss_inquire_cred_by_mech checks its mech argument against two of
 the four mechs a krb5 cred might have (the krb5 mech and the old krb5
 mech, but not the wrong Microsoft mech or the IAKERB mech), so would
 spuriously fail for the other two mechs.  There is no reason to check
 the mechanism if we assume a reasonable mechglue is interpreting
 application gss_inquire_cred_by_mech calls, so just remove the check.

ticket: 7084
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25682 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25660 from trunk
Tom Yu [Fri, 10 Feb 2012 21:19:07 +0000 (21:19 +0000)]
Pull up r25660 from trunk

 ------------------------------------------------------------------------
 r25660 | ghudson | 2012-01-26 16:56:16 -0500 (Thu, 26 Jan 2012) | 22 lines

 ticket: 7082
 subject: Various lookaside cache fixes
 target_version: 1.10
 tags: pullup

 Don't touch the lookaside cache if we're responding with a lookaside
 cache entry.  Also, leave the null entry behind if we're deliberately
 dropping a request (a rare case) so that we don't have to process it
 again.  Fixes several lookaside problems in 1.10:

 * When dropping a request because it was already being processed, we
   were erroneously removing the null entry, causing us to process the
   request again upon a second retransmit.

 * When responding to a finished request with a lookaside entry, we
   were removing and re-adding the entry to the cache, resetting its
   time and performing unnecessary work.

 * We were not caching responses we couldn't deliver because they were
   too big for UDP, causing us to re-process the request when it came
   in again via TCP instead of simply delivering the cached response.

ticket: 7082
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25681 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25659 from trunk
Tom Yu [Fri, 10 Feb 2012 21:19:01 +0000 (21:19 +0000)]
Pull up r25659 from trunk

 ------------------------------------------------------------------------
 r25659 | ghudson | 2012-01-26 12:01:05 -0500 (Thu, 26 Jan 2012) | 10 lines

 ticket: 7081
 subject: Don't use stack variable address in as_req state
 target_version: 1.10
 tags: pullup

 The KDC's process_as_req was storing the address of a stack variable
 (client_princ), which fails if the request is not immediately
 serviced.  Move that variable to the state structure so its address
 remains valid for the lifetime of the request.

ticket: 7081
version_fixed: 1.10.1
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25680 dc483132-0cff-0310-8789-dd5450dbe970

12 years agokrb5-1.10-postrelease
Tom Yu [Fri, 27 Jan 2012 20:16:21 +0000 (20:16 +0000)]
krb5-1.10-postrelease

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25664 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoREADME and patchlevel.h for krb5-1.10
Tom Yu [Fri, 27 Jan 2012 19:42:58 +0000 (19:42 +0000)]
README and patchlevel.h for krb5-1.10

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25662 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoUse INSTALL_DATA to install message catalogues
Tom Yu [Wed, 18 Jan 2012 16:37:02 +0000 (16:37 +0000)]
Use INSTALL_DATA to install message catalogues

Pull up r25492 from trunk

 ------------------------------------------------------------------------
 r25492 | hartmans | 2011-11-29 16:22:26 -0500 (Tue, 29 Nov 2011) | 4 lines

 ticket: new target_version: 1.10 tags: pullup subject: Use INSTALL_DATA to install message catalogues

 Use INSTALL_DATA to avoid marking .mo files executable

ticket: 7078
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25656 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoLIBS should not include PKINIT_CRYPTO_IMPL_LIBS
Tom Yu [Wed, 18 Jan 2012 16:36:57 +0000 (16:36 +0000)]
LIBS should not include PKINIT_CRYPTO_IMPL_LIBS

Pull up r25491 from trunk

 ------------------------------------------------------------------------
 r25491 | hartmans | 2011-11-29 16:22:21 -0500 (Tue, 29 Nov 2011) | 6 lines

 ticket: new Subject: LIBS should not include PKINIT_CRYPTO_IMPL_LIBS tags: pullup target_version: 1.10

 AC_CHECK_LIB should put -lcrypto in PKINIT_CRYPTO_IMPL_LIBS not LIBS
 for pkinit.  A similar problem exists for crypto_impl and is not
 addressed by this patch.

ticket: 7077
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25655 dc483132-0cff-0310-8789-dd5450dbe970

12 years agokrb5-1.10-beta1-postrelease
Tom Yu [Wed, 11 Jan 2012 23:46:42 +0000 (23:46 +0000)]
krb5-1.10-beta1-postrelease

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25647 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoREADME and patchlevel.h for krb5-1.10-beta1
Tom Yu [Wed, 11 Jan 2012 23:32:56 +0000 (23:32 +0000)]
README and patchlevel.h for krb5-1.10-beta1

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25645 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25584 from trunk
Tom Yu [Wed, 11 Jan 2012 03:59:22 +0000 (03:59 +0000)]
Pull up r25584 from trunk

 ------------------------------------------------------------------------
 r25584 | ghudson | 2011-12-12 19:53:56 -0500 (Mon, 12 Dec 2011) | 9 lines

 ticket: 7051
 subject: krb5_server_decrypt_ticket_keytab wrongly succeeds

 If krb5_server_decrypt_ticket_keytab doesn't find a key of the
 appropriate enctype in an iterable keytab, it returns 0 (without
 decrypting the ticket) due to a misplaced initialization of retval.
 This bug causes kinit -k to claim "keytab entry valid" when it
 shouldn't.  Reported by mark@mproehl.net.

ticket: 7051
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25641 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25605 from trunk
Tom Yu [Tue, 10 Jan 2012 21:36:47 +0000 (21:36 +0000)]
Pull up r25605 from trunk

 ------------------------------------------------------------------------
 r25605 | ghudson | 2011-12-27 20:31:59 -0500 (Tue, 27 Dec 2011) | 7 lines

 ticket: 7057

 Fix implicit declaration in ksu for some builds

 ksu's setenv implementation needs to include <string.h> for memcpy.
 Patch from basch@alum.mit.edu.

ticket: 7057
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25639 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25633 from trunk
Tom Yu [Tue, 10 Jan 2012 17:52:37 +0000 (17:52 +0000)]
Pull up r25633 from trunk

 ------------------------------------------------------------------------
 r25633 | tlyu | 2012-01-09 23:23:56 -0500 (Mon, 09 Jan 2012) | 8 lines

 ticket: 7067
 subject: documentation license to CC-BY-SA 3.0 Unported
 target_version: 1.10
 tags: pullup

 Update documentation license to Creative Commons
 Attribution-ShareAlike 3.0 Unported.

ticket: 7067
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25637 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25631 from trunk
Tom Yu [Tue, 10 Jan 2012 17:43:23 +0000 (17:43 +0000)]
Pull up r25631 from trunk

 ------------------------------------------------------------------------
 r25631 | tlyu | 2012-01-09 16:35:30 -0500 (Mon, 09 Jan 2012) | 7 lines

 ticket: 7065
 subject: delete duplicate NOTICE file
 status: open

 Delete the duplicate NOTICE file, as it is checked into the tree and
 relative include paths work for sphinx.

ticket: 7065
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25636 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25621 from trunk
Tom Yu [Mon, 9 Jan 2012 21:23:48 +0000 (21:23 +0000)]
Pull up r25621 from trunk

 ------------------------------------------------------------------------
 r25621 | ghudson | 2012-01-07 15:57:36 -0500 (Sat, 07 Jan 2012) | 9 lines

 ticket: 7060
 subject: Convert securid module edata method
 target_version: 1.10
 tags: pullup

 r25348 made modified the edata method of the kdcpreauth interface to
 be async-capable, but neglected to convert the securid_sam2 module's
 edata function.  Do that now.

ticket: 7060
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25630 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25598 from trunk
Tom Yu [Mon, 9 Jan 2012 21:23:46 +0000 (21:23 +0000)]
Pull up r25598 from trunk

 ------------------------------------------------------------------------
 r25598 | ghudson | 2011-12-20 13:08:07 -0500 (Tue, 20 Dec 2011) | 7 lines

 Table of Contents.hhc -> Table_of_Contents.hhc

 Avoid using spaces in filenames as it makes searching the source
 tree less convenient on Unix.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
ticket: 7055
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25629 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25591, r25604 from trunk
Tom Yu [Mon, 9 Jan 2012 21:23:43 +0000 (21:23 +0000)]
Pull up r25591, r25604 from trunk

 ------------------------------------------------------------------------
 r25604 | ghudson | 2011-12-27 01:39:07 -0500 (Tue, 27 Dec 2011) | 4 lines

 ticket: 6936

 Fix an unlikely memory leak in r25591
 ------------------------------------------------------------------------
 r25591 | ghudson | 2011-12-16 18:19:01 -0500 (Fri, 16 Dec 2011) | 14 lines

 ticket: 6936
 target_version: 1.10
 tags: pullup

 Do mech fallback for first SPNEGO context token

 When producing the first SPNEGO security context token, if the first
 mechanism's init_sec_context fails, fall back to a later mechanism.

 This fixes a regression in 1.10 for SPNEGO initiators using non-krb5
 credentials.  The identity selection work causes errors to be deferred
 from krb5's acquire_cred in some cases, which means SPNEGO doesn't see
 an error until it tries the krb5 init_sec_context.

ticket: 6936
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25628 dc483132-0cff-0310-8789-dd5450dbe970

12 years agoPull up r25590 from trunk
Tom Yu [Mon, 9 Jan 2012 21:23:40 +0000 (21:23 +0000)]
Pull up r25590 from trunk

 ------------------------------------------------------------------------
 r25590 | ghudson | 2011-12-16 18:18:54 -0500 (Fri, 16 Dec 2011) | 11 lines

 ticket: 7053
 subject: Verify acceptor's mech in SPNEGO initiator
 target_version: 1.10
 tags: pullup

 In spnego_gss_ctx_id_rec, store the set of negotiable mechanisms as
 well as the currently selected internal_mech, which becomes an alias
 into mech_set.  In init_ctx_reselect, locate the acceptor's counter-
 proposal in sc->mech_set and consider the token defective if it is not
 found.

ticket: 7053
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25627 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoUpdate acknowledgments and some summary info
Tom Yu [Thu, 22 Dec 2011 21:27:16 +0000 (21:27 +0000)]
Update acknowledgments and some summary info

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25603 dc483132-0cff-0310-8789-dd5450dbe970

13 years agokrb5-1.10-alpha2-postrelease
Tom Yu [Tue, 20 Dec 2011 04:01:28 +0000 (04:01 +0000)]
krb5-1.10-alpha2-postrelease

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25597 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoREADME and patchlevel.h for krb5-1.10-alpha2
Tom Yu [Mon, 19 Dec 2011 21:21:58 +0000 (21:21 +0000)]
README and patchlevel.h for krb5-1.10-alpha2

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25595 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25536 from trunk
Tom Yu [Wed, 14 Dec 2011 23:32:41 +0000 (23:32 +0000)]
pull up r25536 from trunk

 ------------------------------------------------------------------------
 r25536 | ghudson | 2011-12-09 12:57:52 -0500 (Fri, 09 Dec 2011) | 8 lines

 ticket: 7049
 subject: Fix subkey memory leak in krb5_get_credentials
 target_version: 1.10
 tags: pullup

 If a get_credentials operation requires multiple TGS requests, we need
 to free the subkey from previous requests before saving a new one.

ticket: 7049
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25586 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoSquash commits for KfW updates
Tom Yu [Wed, 14 Dec 2011 00:07:54 +0000 (00:07 +0000)]
Squash commits for KfW updates

windows ccapiserver: replace Sleep with event wait

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
fix warning in test_cc_credentials_iterator_next.c

include test_ccapi_iterators.h for check_cc_credentials_iterator_next

Make ccapiserver exit if its receiveloop thread terminates for any reason.

This happens, for example, when the rpc endpoint is already registered
by another ccapiserver process.  There's no reason to leave a zombie
process running that can't receive messages.

windows ccapi: launch server without console by default.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
windows ccapi: use a random challenge to authenticate ccapiserver.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
LeashView.cpp: only specify TVIF_TEXT if there is actually text.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: add runtime.wxi WIXINCLUDES in Makefile to fix dependencies.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
Windows leash64 fixes: use proper names for leash and krb5 dlls

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
Windows leash fixes: 'make install' installs leash exes.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: use MSVC 2010 merge modules

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: install leash32.exe

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: clean out unused #defines from Lglobals.h

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: use correct message id to obtain tgt from leash

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: update copyright notice in license.rtf

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fixes: install xpprof32

TODO: xpprof64!

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: purge support for old compilers

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: don't build installer into installer

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fixes: make leash ignore credentials that store config principals.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fix: make Leash_kdestroy() actually destroy k5 tickets

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fix: Add custom "Password incorrect" message to Leash_int_kinit_ex()

Overrides obscure KRB5KRB_AP_ERR_BAD_INTEGRITY message.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fixes: define USE_MESSAGE_BOX in leashdll code for user feedback.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fixes: krb5_get_init_creds_opt_init->krb5_get_init_creds_opt_alloc

Should enable leash to generate config credentials (needs verification!)

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fix: int -> size_t to fix warning in krb5routines.c

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fix: restructure low ticket warning popup code to workaround mfc bug

mfc bug causes assertions when dialog is generated from
within PreTranslateMessages() (MSG input param points to a global
variable which is corrupted in the dialog message loop).  So we need
to instead PostMessage() to cause the popup later.
Also fixed logic to cause warning dialog to actually be modal as intended
when the leash window is not minimized.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fixes: fix _snprintf usage; use full error code in leash_error_message

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw fixes: ccapiserver only quits after all clients detach.

Not sure if this is really a good idea or not...

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: generate manifests

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: generate leash shortcuts (desktop and start menu)

...also install xpprof64

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: fix 'K5_ORIGINAL_NAME' for 64 bit dlls.

...still need to actually to define _WIN64 for rc.exe though

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: purge bufferoverflowu from custom.dll

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: rename leash32/64.exe to simply leash.exe

Also install leash.exe in 64 bit installer.

Split cci_thread_init into per-process and per-thread portions

Call the per-thread code on thread attach and per-process once per
process.  Previously, while the function was named 'thread', it was
only actually called once per process.  Currently, the per-thread
code does nothing on non-windows platforms and is not even actually
invoked.

Fixes a windows bug when multiple non-main threads try to use ccapi
at the same time.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw leash: add -console option to create console for debug output

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: use _WIN64 names where appropriate

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw leash: bracket krb.con code with #ifndef NO_KRB4

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: install krb5.ini to CommonAppDataFolder, not WindowsFolder

...but only if there isn't already a krb5.ini in the WindowsFolder.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: "make install" also installs pdbs

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: leash32.pdb->leash.pdb

kfw installer: add site-local.wxi

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: leash htmlhelp file source

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: use html help in leash

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: "make install" installs htmlhelp (leash.chm)

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw installer: install leash help file (leash.chm)

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw: remove line breaks from html to fix table of contents generation

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw leash help: fix/add aliases for command help

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
kfw leash: fix bad data in get tickets dialog when -autoinit specified

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
ticket: 7050
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25585 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoMake depend
Tom Yu [Wed, 7 Dec 2011 05:26:47 +0000 (05:26 +0000)]
Make depend

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25528 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25525 from trunk
Tom Yu [Tue, 6 Dec 2011 20:52:18 +0000 (20:52 +0000)]
pull up r25525 from trunk

 ------------------------------------------------------------------------
 r25525 | tlyu | 2011-12-06 15:42:46 -0500 (Tue, 06 Dec 2011) | 8 lines

 ticket: 7042
 subject: SA-2011-007 KDC null pointer deref in TGS handling [CVE-2011-1530]
 target_version: 1.10
 tags: pullup

 Fix a null pointer dereference condition that could cause a denial of
 service.

ticket: 7042
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25526 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25504 from trunk
Tom Yu [Tue, 6 Dec 2011 00:07:11 +0000 (00:07 +0000)]
pull up r25504 from trunk

 ------------------------------------------------------------------------
 r25504 | ghudson | 2011-12-04 17:38:36 -0500 (Sun, 04 Dec 2011) | 20 lines

 ticket: 7033
 target_version: 1.10
 tags: pullup

 Set a default enctype for optimistic preauth

 When the client application requests optimistic preauth for a preauth
 type which uses the password, we don't have an etype-info2 to
 interpret since we haven't talked to the KDC.  So we need to guess an
 enctype, salt, and s2k parameters.  In 1.9 and prior, encrypted
 timestamp contained code to use the first requested enctype in this
 case, but encrypted challenge did not.  In 1.10 prior to this change,
 neither mechanism uses a reasonable default.

 Set a default enctype in krb5_init_creds_init so that all
 password-based preauth mechanisms will use a reasonable default in the
 optimistic preauth case.  The default salt and s2k parameters for this
 case will be the principal-based default salt and the enctype-based
 default parameters.

ticket: 7033
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25524 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25500 from trunk
Tom Yu [Mon, 5 Dec 2011 23:43:31 +0000 (23:43 +0000)]
pull up r25500 from trunk

 ------------------------------------------------------------------------
 r25500 | hartmans | 2011-12-02 14:42:12 -0500 (Fri, 02 Dec 2011) | 7 lines

 ticket: 7038
 subject: Added support for loading of Krb5.ini from Windows APPDATA
 target_version: 1.10
 tags: pullup

Signed-off-by: Alexey Melnikov <aamelnikov@gmail.com>
ticket: 7038
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25523 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25499 from trunk
Tom Yu [Mon, 5 Dec 2011 23:43:28 +0000 (23:43 +0000)]
pull up r25499 from trunk

 ------------------------------------------------------------------------
 r25499 | hartmans | 2011-12-02 14:20:54 -0500 (Fri, 02 Dec 2011) | 7 lines

 ticket: 7037
 subject:  Use LsaDeregisterLogonProcess(), not CloseHandle()
 target_version: 1.10
 tags: pullup

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
ticket: 7037
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25522 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25498 from trunk
Tom Yu [Mon, 5 Dec 2011 23:43:25 +0000 (23:43 +0000)]
pull up r25498 from trunk

 ------------------------------------------------------------------------
 r25498 | hartmans | 2011-12-02 13:52:22 -0500 (Fri, 02 Dec 2011) | 8 lines

 ticket: 7036
 subject: Fix free ofuninitialized memory in sname_to_princ
 tags: pullup
 Target_Version: 1.10

 Fix free of uninitialized memory in error case introduced in 1.10
 development cycle.

ticket: 7036
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25521 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25497 from trunk
Tom Yu [Mon, 5 Dec 2011 23:43:23 +0000 (23:43 +0000)]
pull up r25497 from trunk

 ------------------------------------------------------------------------
 r25497 | hartmans | 2011-12-02 13:52:19 -0500 (Fri, 02 Dec 2011) | 7 lines

 ticket: 7035
 subject:  krb5_lcc_store() now ignores config credentials
 target_version: 1.10
 tags: pullup

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
ticket: 7035
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25520 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25496 from trunk
Tom Yu [Mon, 5 Dec 2011 23:43:17 +0000 (23:43 +0000)]
pull up r25496 from trunk

 ------------------------------------------------------------------------
 r25496 | hartmans | 2011-12-02 13:52:12 -0500 (Fri, 02 Dec 2011) | 11 lines

 ticket: 7034
 subject: mk_cred: memory management
 target_version: 1.10
 tags: pullup

 Fix for mk_cred.c: calloc() not malloc()

 Avoid calling free() in cleanup on uninitialized sub-ptrs if error occurs.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
ticket: 7034
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25519 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25494 from trunk
Tom Yu [Mon, 5 Dec 2011 23:43:14 +0000 (23:43 +0000)]
pull up r25494 from trunk

 ------------------------------------------------------------------------
 r25494 | hartmans | 2011-11-29 18:11:13 -0500 (Tue, 29 Nov 2011) | 7 lines

 ticket: 7030
 subject: Ldap dependency for parallel builds
 tags: pullup
 target_version: 1.10

 The ldap plugin needs to declare a dependency on the ldap library

ticket: 7030
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25518 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25493 from trunk
Tom Yu [Mon, 5 Dec 2011 23:43:11 +0000 (23:43 +0000)]
pull up r25493 from trunk

 ------------------------------------------------------------------------
 r25493 | ghudson | 2011-11-29 17:49:56 -0500 (Tue, 29 Nov 2011) | 9 lines

 ticket: 7029
 subject: Fix --with-system-verto without pkg-config
 target_version: 1.10
 tags: pullup

 If we're using the system verto and pkg-config isn't found but
 libverto is, set VERTO_LIBS to just -lverto as there won't be a k5ev
 module.

ticket: 7029
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25517 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25486 from trunk
Tom Yu [Mon, 5 Dec 2011 23:12:03 +0000 (23:12 +0000)]
pull up r25486 from trunk

 ------------------------------------------------------------------------
 r25486 | hartmans | 2011-11-22 20:00:27 -0500 (Tue, 22 Nov 2011) | 14 lines

     ticket: new
     subject: FAST PKINIT
     target_version: 1.10
     tags: pullup

     Per RFC 6113 fast should use the inner request body for the pkinit
     checksum. We did that on the KDC; now do so on the client.  Remove
     code that explicitly blocked pkinit under FAST.

     Also, use the reply key *before* the strengthen key is applied when
     verifying the PADATA_PKINIT_KX.

     Add FAST pkinit test.

ticket: 7027
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25516 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25483 and r25484 from trunk
Tom Yu [Mon, 5 Dec 2011 23:11:57 +0000 (23:11 +0000)]
pull up r25483 and r25484 from trunk

 ------------------------------------------------------------------------
 r25484 | ghudson | 2011-11-22 12:48:29 -0500 (Tue, 22 Nov 2011) | 7 lines

 ticket: 7023

 Fix compile error in previous change

 A last-minute code editing mistake crept into the previous commit; fix
 it.

 ------------------------------------------------------------------------
 r25483 | ghudson | 2011-11-21 16:14:39 -0500 (Mon, 21 Nov 2011) | 21 lines

 ticket: 7023
 subject: Clean up client-side preauth error data handling
 target_version: 1.10
 tags: pullup

 Change the clpreauth tryagain method to accept a list of pa-data,
 taken either from the FAST response or from decoding the e_data as
 either pa-data or typed-data.  Also change the in_padata argument to
 contain just the type of the request padata rather than the whole
 element, since modules generally shouldn't care about the contents of
 their request padata (or they can remember it).

 In krb5int_fast_process_error, no longer re-encode FAST pa-data as
 typed-data for the inner error e_data, but decode traditional error
 e_data for all error types, and try both pa-data and typed-data
 encoding.

 In PKINIT, try all elements of the new pa-data list, since it may
 contain FAST elements as well as the actual PKINIT array.  (Fixes an
 outstanding bug in FAST PKINIT.)

ticket: 7023
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25515 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25480 from trunk
Tom Yu [Mon, 5 Dec 2011 20:53:52 +0000 (20:53 +0000)]
pull up r25480 from trunk

 ------------------------------------------------------------------------
 r25480 | ghudson | 2011-11-20 00:19:45 -0500 (Sun, 20 Nov 2011) | 13 lines

 ticket: 7021
 subject: Fix failure interval of 0 in LDAP lockout code
 target_version: 1.10
 tags: pullup

 A failure count interval of 0 caused krb5_ldap_lockout_check_policy to
 pass the lockout check (but didn't cause a reset of the failure count
 in krb5_ldap_lockout_audit).  It should be treated as forever, as in
 the DB2 back end.

 This bug is the previously unknown cause of the assertion failure
 fixed in CVE-2011-1528.

ticket: 7021
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25512 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25482 from trunk
Tom Yu [Mon, 5 Dec 2011 20:32:57 +0000 (20:32 +0000)]
pull up r25482 from trunk

 ------------------------------------------------------------------------
 r25482 | ghudson | 2011-11-21 12:30:41 -0500 (Mon, 21 Nov 2011) | 10 lines

 ticket: 7020
 target_version: 1.10
 tags: pullup

 Recognize IAKERB mech in krb5_gss_display_status

 Minor status codes were not displaying properly when originated from
 the IAKERB mech, because of a safety check on mech_type.  From Ralf
 Haferkamp <rhafer@suse.de>.

ticket: 7020
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25511 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25475 and r25479 from trunk
Tom Yu [Mon, 5 Dec 2011 20:16:05 +0000 (20:16 +0000)]
pull up r25475 and r25479 from trunk

 ------------------------------------------------------------------------
 r25479 | ghudson | 2011-11-19 17:06:15 -0500 (Sat, 19 Nov 2011) | 8 lines

 ticket: 7019

 Improve documentation in preauth_plugin.h

 Also declare the verto_context structure to ensure that it is has the
 proper scope when used as the return type of the event_context
 callback.

 ------------------------------------------------------------------------
 r25475 | ghudson | 2011-11-14 21:42:58 -0500 (Mon, 14 Nov 2011) | 9 lines

 ticket: 7019
 subject: Make verto context available to kdcpreauth modules
 target_version: 1.10
 tags: pullup

 Add an event_context callback to kdcpreauth.  Adjust the internal KDC
 and main loop interfaces to pass around the event context, and expose
 it to kdcpreauth modules via the rock.

ticket: 7019
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25510 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25474 from trunk
Tom Yu [Mon, 5 Dec 2011 19:44:11 +0000 (19:44 +0000)]
pull up r25474 from trunk

 ------------------------------------------------------------------------
 r25474 | ghudson | 2011-11-14 20:59:01 -0500 (Mon, 14 Nov 2011) | 10 lines

 ticket: 7018
 subject: Update verto to 0.2.2 release
 target_version: 1.10
 tags: pullup

 Update verto sources to 0.2.2 release versions.  verto_reinitialize()
 now has a return value; check it in kdc/main.c.  Store verto-libev.c
 alongside verto-k5ev.c to make it easy to diff corresponding versions
 when updating.

ticket: 7018
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25509 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25473 from trunk
Tom Yu [Mon, 5 Dec 2011 19:44:05 +0000 (19:44 +0000)]
pull up r25473 from trunk

 ------------------------------------------------------------------------
 r25473 | ghudson | 2011-11-14 16:45:33 -0500 (Mon, 14 Nov 2011) | 16 lines

 ticket: 7017
 subject: Simplify and fix kdcpreauth request_body callback
 target_version: 1.10
 tags: pullup

 Alter the contract for the kdcpreauth request_body callback so that it
 returns an alias to the encoded body instead of a fresh copy.  At the
 beginning of AS request processing, save a copy of the encoded request
 body, or the encoded inner request body for FAST requests.  Previously
 the request_body callback would re-encode the request structure, which
 in some cases has been modified by the AS request code.

 No kdcpreauth modules currently use the request_body callback, but
 PKINIT will need to start using it in order to handle FAST requests
 correctly.

ticket: 7017
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25508 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoHandle TGS referrals to the same realm
Tom Yu [Mon, 5 Dec 2011 19:01:49 +0000 (19:01 +0000)]
Handle TGS referrals to the same realm

pull up r25472 from trunk

 ------------------------------------------------------------------------
 r25472 | ghudson | 2011-11-14 13:02:52 -0500 (Mon, 14 Nov 2011) | 12 lines

 ticket: 7016
 subject: Handle TGS referrals to the same realm
 target_version: 1.9.3
 tags: pullup

 krb5 1.6 through 1.8 contained a workaround for the Active Directory
 behavior of returning a TGS referral to the same realm as the request.
 1.9 responds to this behavior by caching the returned TGT, trying
 again, and detecting a referral loop.  This is a partial regression of
 ticket #4955.  Detect this case and fall back to a non-referreal
 request.

ticket: 7039
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25507 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25470 from trunk
Tom Yu [Fri, 2 Dec 2011 21:20:22 +0000 (21:20 +0000)]
pull up r25470 from trunk

 ------------------------------------------------------------------------
 r25470 | ghudson | 2011-11-12 17:03:54 -0500 (Sat, 12 Nov 2011) | 9 lines

 ticket: 7015
 subject: Add plugin interface_names entry for ccselect
 target_version: 1.10
 tags: pullup

 When the ccselect pluggable interface was added, the interface_names
 table wasn't updated, so configuring modules for it wouldn't work.
 Add it now.

ticket: 7015
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25503 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25468 from trunk
Tom Yu [Fri, 2 Dec 2011 21:20:18 +0000 (21:20 +0000)]
pull up r25468 from trunk

 ------------------------------------------------------------------------
 r25468 | ghudson | 2011-11-10 23:04:58 -0500 (Thu, 10 Nov 2011) | 12 lines

 ticket: 7014
 subject: Fix com_err.h dependencies in gss-kernel-lib
 target_version: 1.10
 tags: pullup

 make check was failing in util/gss-kernel-lib due to dependencies
 when the build is configured with --with-system-et, because depfix.pl
 wasn't smart enough to substitute the dependency on com_err.h in the
 current directory.  Make depfix.pl smarter, and adjust COM_ERR_DEPS
 to be com_err.h in gss-kernel-lib when building with the bundled
 com_err.

ticket: 7014
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25502 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25469 from trunk
Tom Yu [Fri, 2 Dec 2011 21:20:13 +0000 (21:20 +0000)]
pull up r25469 from trunk

 ------------------------------------------------------------------------
 r25469 | ghudson | 2011-11-11 12:01:12 -0500 (Fri, 11 Nov 2011) | 14 lines

 ticket: 6430
 subject: Avoid looping when preauth can't be generated
 target_version: 1.10
 tags: pullup

 If we receive a PREAUTH_REQUIRED error and fail to generate any real
 preauthentication, error out immediately instead of continuing to
 generate non-preauthenticated requests until we hit the loop count.

 There is a lot of room to generate a more meaningful error about why
 we failed to generate preauth (although in many cases the answer may
 be too complicated to explain in an error message), but that requires
 more radical restructuring of the preauth framework.

ticket: 6430
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25501 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25424 from trunk
Tom Yu [Mon, 7 Nov 2011 22:51:36 +0000 (22:51 +0000)]
pull up r25424 from trunk

 ------------------------------------------------------------------------
 r25424 | ghudson | 2011-10-31 12:43:40 -0400 (Mon, 31 Oct 2011) | 9 lines

 ticket: 6996
 subject: Make krb5_check_clockskew public
 target_version: 1.10
 tags: pullup

 Rename krb5int_check_clockskew to krb5_check_clockskew and make it
 public, in order to give kdcpreauth plugins a way to check timestamps
 against the configured clock skew.

ticket: 6996
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25456 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25444 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:54 +0000 (22:35 +0000)]
pull up r25444 from trunk

 ------------------------------------------------------------------------
 r25444 | ghudson | 2011-11-06 00:32:34 -0500 (Sun, 06 Nov 2011) | 10 lines

 ticket: 7003
 subject: Fix month/year units in getdate
 target_version: 1.10
 tags: pullup

 getdate strings like "1 month" or "next year" would fail some of the
 time, depending on the value of stack garbage, because DSTcorrect()
 doesn't set *error on success and RelativeMonth() doesn't initialize
 error.  Make DSTcorrect() responsible for setting *error in all cases.

ticket: 7003
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25455 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25443 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:51 +0000 (22:35 +0000)]
pull up r25443 from trunk

 ------------------------------------------------------------------------
 r25443 | ghudson | 2011-11-05 15:55:34 -0400 (Sat, 05 Nov 2011) | 11 lines

 ticket: 7002
 target_version: 1.10
 tags: pullup

 Improve verto and libev documentation

 NOTICE was missing the copyright statement for verto (it's not quite
 the same as other Red Hat licenses).  util/verto had no README file,
 and neither the verto nor k5ev README contained pointers to the
 upstream project pages.

ticket: 7002
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25454 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25433 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:48 +0000 (22:35 +0000)]
pull up r25433 from trunk

 ------------------------------------------------------------------------
 r25433 | ghudson | 2011-11-04 01:53:23 -0400 (Fri, 04 Nov 2011) | 9 lines

 ticket: 7000
 subject: Exit on error in kadmind kprop child
 target_version: 1.10
 tags: pullup

 When we fork from kadmind to dump the database and kprop to an iprop
 slave, if we encounter an error in the child process we should exit
 rather than returning to the main loop.

ticket: 7000
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25453 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25445 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:45 +0000 (22:35 +0000)]
pull up r25445 from trunk

 ------------------------------------------------------------------------
 r25445 | ghudson | 2011-11-06 19:47:20 -0500 (Sun, 06 Nov 2011) | 8 lines

 ticket: 6999
 target_version: 1.10
 tags: pullup

 Fix warnings and version check for NSS pkinit

 From nalin@redhat.com.

ticket: 6999
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25452 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25425 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:41 +0000 (22:35 +0000)]
pull up r25425 from trunk

 ------------------------------------------------------------------------
 r25425 | ghudson | 2011-10-31 23:49:16 -0400 (Mon, 31 Oct 2011) | 10 lines

 ticket: 6997
 target_version: 1.10
 tags: pullup

 Conditionalize po subdir on msgfmt, not dgetext

 The presence of dgettext in libc or libintl doesn't imply that msgfmt
 is installed, so conditionalize building the po subdir on whether
 msgfmt is installed.

ticket: 6997
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25451 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25419 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:39 +0000 (22:35 +0000)]
pull up r25419 from trunk

 ------------------------------------------------------------------------
 r25419 | ghudson | 2011-10-28 11:53:50 -0400 (Fri, 28 Oct 2011) | 11 lines

 ticket: 6995
 subject: Initialize typed_e_data in as_req_state
 target_version: 1.10
 tags: pullup

 The typed_e_data field in struct as_req_state was not properly
 initialized, causing the KDC to sometimes respond with typed-data
 e_data for a preauth-required error when the client sends no padata.
 This bug was masked with recent clients, which send a
 KRB5_ENCPADATA_REQ_ENC_PA_REP padata.

ticket: 6995
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25450 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25418 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:35 +0000 (22:35 +0000)]
pull up r25418 from trunk

 ------------------------------------------------------------------------
 r25418 | ghudson | 2011-10-28 11:45:03 -0400 (Fri, 28 Oct 2011) | 9 lines

 ticket: 6994
 subject: Fix intermediate key length in hmac-md5 checksum
 target_version: 1.10
 tags: pullup

 When using hmac-md5, the intermediate key length is the output of the
 hash function (128 bits), not the input key length.  Relevant if the
 input key is not an RC4 key.

ticket: 6994
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25449 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoFix format string for TRACE_INIT_CREDS_SERVICE
Tom Yu [Mon, 7 Nov 2011 22:35:32 +0000 (22:35 +0000)]
Fix format string for TRACE_INIT_CREDS_SERVICE

pull up r25417 from trunk

 ------------------------------------------------------------------------
 r25417 | ghudson | 2011-10-26 18:34:21 -0400 (Wed, 26 Oct 2011) | 7 lines

 ticket: 6993
 subject: Fix format string for TRACE_INIT_CREDS_SERVICE
 tags: pullup
 target_version: 1.9.2

 This should also be pulled up to 1.10.

ticket: 7006
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25448 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25414 from trunk
Tom Yu [Mon, 7 Nov 2011 22:35:24 +0000 (22:35 +0000)]
pull up r25414 from trunk

 ------------------------------------------------------------------------
 r25414 | ghudson | 2011-10-25 14:30:14 -0400 (Tue, 25 Oct 2011) | 7 lines

 ticket: 6992
 subject: Make krb5_find_authdata public
 target_version: 1.10
 tags: pullup

 Rename krb5int_find_authdata to krb5_find_authdata and make it public.

ticket: 6992
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25447 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoUpdate acknowledgments
Tom Yu [Sat, 5 Nov 2011 00:03:44 +0000 (00:03 +0000)]
Update acknowledgments

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25442 dc483132-0cff-0310-8789-dd5450dbe970

13 years agokrb5-1.10-alpha1-postrelease
Tom Yu [Fri, 21 Oct 2011 18:03:54 +0000 (18:03 +0000)]
krb5-1.10-alpha1-postrelease

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25402 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoREADME and patchlevel.h for krb5-1.10-alpha1
Tom Yu [Fri, 21 Oct 2011 18:02:10 +0000 (18:02 +0000)]
README and patchlevel.h for krb5-1.10-alpha1

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25400 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25395 from trunk
Tom Yu [Fri, 21 Oct 2011 17:58:47 +0000 (17:58 +0000)]
pull up r25395 from trunk

 ------------------------------------------------------------------------
 r25395 | tlyu | 2011-10-21 13:35:49 -0400 (Fri, 21 Oct 2011) | 10 lines

 ticket: 6989
 subject: fix tar invocation in mkrel
 target_version: 1.10
 tags: pullup

 Fix the tar invocation in mkrel so that it defaults to using "tar" as
 the tar program rather than "gtar".

 This should probably be pulled up to at least 1.9 and 1.8 as well.

ticket: 6989
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25399 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoFix patchlevel.h for krb5-1.10 branch
Tom Yu [Thu, 20 Oct 2011 22:19:39 +0000 (22:19 +0000)]
Fix patchlevel.h for krb5-1.10 branch

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25392 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoUpdate README for 1.10 branch
Tom Yu [Thu, 20 Oct 2011 22:13:09 +0000 (22:13 +0000)]
Update README for 1.10 branch

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25391 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25385 from trunk
Tom Yu [Thu, 20 Oct 2011 19:27:46 +0000 (19:27 +0000)]
pull up r25385 from trunk

 ------------------------------------------------------------------------
 r25385 | ghudson | 2011-10-20 11:16:03 -0400 (Thu, 20 Oct 2011) | 9 lines

 ticket: 6988
 subject: Fix handling of null edata method in KDC preauth
 target_version: 1.10
 tags: pullup

 Correctly include an empty padata value if a KDC preauth system has no
 get_edata method.  This bug prevented the KDC from indicating FAST
 support in preauth-required errors.

ticket: 6988
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25389 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25384 from trunk
Tom Yu [Thu, 20 Oct 2011 19:27:43 +0000 (19:27 +0000)]
pull up r25384 from trunk

 ------------------------------------------------------------------------
 r25384 | ghudson | 2011-10-19 23:45:12 -0400 (Wed, 19 Oct 2011) | 12 lines

 ticket: 6987
 subject: Fix krb5_cc_set_config
 target_version: 1.10
 tags: pullup

 krb5_cc_set_config has been non-functional since r24753 on cache types
 which don't support removal of credential entries.  Fix it by only
 calling krb5_cc_remove_cred if data is NULL, since krb5_cc_store_cred
 will do it anyway in the positive case.

 Also fix an old memory leak in an uncommon error case.

ticket: 6987
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25388 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25368 from trunk
Tom Yu [Thu, 20 Oct 2011 19:27:38 +0000 (19:27 +0000)]
pull up r25368 from trunk

 ------------------------------------------------------------------------
 r25368 | tlyu | 2011-10-18 14:51:35 -0400 (Tue, 18 Oct 2011) | 8 lines

 ticket: 6981
 subject: SA-2011-006 KDC denial of service [CVE-2011-1527 CVE-2011-1528 CVE-2011-1529]
 target_version: 1.10
 tags: pullup

 Fix null pointer dereference and assertion failure conditions that
 could cause a denial of service.

ticket: 6981
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25387 dc483132-0cff-0310-8789-dd5450dbe970

13 years agopull up r25367 from trunk
Tom Yu [Thu, 20 Oct 2011 19:27:32 +0000 (19:27 +0000)]
pull up r25367 from trunk

 ------------------------------------------------------------------------
 r25367 | ghudson | 2011-10-18 12:32:28 -0400 (Tue, 18 Oct 2011) | 12 lines

 ticket: 6980
 subject: Ensure termination in Windows vsnprintf wrapper
 target_version: 1.10
 tags: pullup

 The Windows _vsnprintf does not terminate its output buffer in the
 overflow case.  Make sure we do that in the wrapper.  Reported by
 Chris Hecker.

 (Not an issue for KfW 3.2 since we weren't using snprintf in 1.6.x
 except in Unix-specific code.)

ticket: 6980
version_fixed: 1.10
status: resolved

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25386 dc483132-0cff-0310-8789-dd5450dbe970

13 years agobranch for krb5-1.10 release
Tom Yu [Mon, 17 Oct 2011 22:55:44 +0000 (22:55 +0000)]
branch for krb5-1.10 release

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-10@25366 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoBump release numbers in definitions.texinfo
Tom Yu [Mon, 17 Oct 2011 22:54:12 +0000 (22:54 +0000)]
Bump release numbers in definitions.texinfo

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25365 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoNoted that kadmind should be restarted if its acl file has been changed
Zhanna Tsitkov [Mon, 17 Oct 2011 20:17:08 +0000 (20:17 +0000)]
Noted that kadmind should be restarted if its acl file has been changed

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25364 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoDelete Network Identity Manager
Tom Yu [Mon, 17 Oct 2011 19:34:08 +0000 (19:34 +0000)]
Delete Network Identity Manager

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25363 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoMake reindent
Tom Yu [Mon, 17 Oct 2011 19:11:01 +0000 (19:11 +0000)]
Make reindent

Also fix pkinit_crypto_nss.c struct initializers and add parens to a
ternary operator in do_as_req.c for better indentation.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25362 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoExclude util/wshelper from reindent
Tom Yu [Mon, 17 Oct 2011 19:10:52 +0000 (19:10 +0000)]
Exclude util/wshelper from reindent

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25361 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoAdd AC_LANG_SOURCE to PKINIT NSS version check
Greg Hudson [Mon, 17 Oct 2011 17:15:31 +0000 (17:15 +0000)]
Add AC_LANG_SOURCE to PKINIT NSS version check

The configure.in code for the PKINIT NSS back end version check was
copied from the k5crypto NSS back end version check, but from before
r25181 which added AC_LANG_SOURCE wrappers.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25360 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoStyle police
Greg Hudson [Mon, 17 Oct 2011 04:05:56 +0000 (04:05 +0000)]
Style police

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25359 dc483132-0cff-0310-8789-dd5450dbe970

13 years agogssalloc-related fixes to naming_exts.c
Sam Hartman [Mon, 17 Oct 2011 00:45:30 +0000 (00:45 +0000)]
gssalloc-related fixes to naming_exts.c

renamed kg_data_list_to_buffer_set_nocopy to data_list_buffer_set
(since nocopy is no longer guaranteed).
removed extra indirection to input krb5_data list.
ensured input krb5_data list is always completely freed.
no longer returns EINVAL when output buffer set is NULL.
fixed krb5_gss_get_name_attribute to use data_to_gss.

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25358 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoRFC 4120 says that we should not canonicalize using DNS. We cannot get
Sam Hartman [Mon, 17 Oct 2011 00:45:23 +0000 (00:45 +0000)]
RFC 4120 says that we should not canonicalize using DNS. We cannot get
that far today, but there's no reason we should fail to use a
perfectly good principal name just because DNS is failing. For some
services there isn't even a requirement they be in DNS. With
AI_ADDRCONFIG there's no reason that Kerberos canonicalization should
fail simply because a v6 address is not present, for example.  So, if
getaddrinfo fails in krb5_sname_to_principal simply use the input
hostname uncanonicalized.

sn2princ: On getaddrinfo failure use the input

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25357 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoAllow password changes over NATs
Greg Hudson [Sat, 15 Oct 2011 16:56:30 +0000 (16:56 +0000)]
Allow password changes over NATs

In the kpasswd server code, don't set a remote address in the auth
context before calling krb5_rd_priv, since the kpasswd protocol is
well-protected against reflection attacks.  This allows password
changes to work in cases where a NAT has changed the client IP address
as it is seen by the server.

ticket: 6979

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25356 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoAllow rd_priv/rd_safe without remote address
Greg Hudson [Sat, 15 Oct 2011 16:56:26 +0000 (16:56 +0000)]
Allow rd_priv/rd_safe without remote address

Allow krb5_rd_priv and krb5_rd_safe to work when there is no remote
address set in the auth context, unless the KRB5_AUTH_CONTEXT_DO_TIMES
flag is set (in which case we need the remote address for the replay
cache name).  Note that failing to set the remote address can create a
vulnerability to reflection attacks in some protocols, although it is
fairly easy to defend against--either use sequence numbers, or make
sure that requests don't look like replies, or both.

ticket: 6978

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25355 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoUpdate mit-krb5.pot
Greg Hudson [Sat, 15 Oct 2011 16:31:00 +0000 (16:31 +0000)]
Update mit-krb5.pot

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25354 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoInstall krb5/preauth_plugin.h
Greg Hudson [Sat, 15 Oct 2011 16:29:28 +0000 (16:29 +0000)]
Install krb5/preauth_plugin.h

The clpreauth and kdcpreauth pluggable interfaces are public as of
krb5 1.10.  Install the header so that preauth modules can be built
outside of the krb5 source tree.

ticket: 6977

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25353 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoRename PAC type constants to avoid conflicts
Greg Hudson [Sat, 15 Oct 2011 16:26:27 +0000 (16:26 +0000)]
Rename PAC type constants to avoid conflicts

Since the PAC type constants are now exposed in krb5.h, give them a
KRB5_ prefix so they don't conflict with similar PAC type constants
in other packages, like Samba.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25352 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoHide gak_fct interface and arguments in clpreauth
Greg Hudson [Sat, 15 Oct 2011 16:06:03 +0000 (16:06 +0000)]
Hide gak_fct interface and arguments in clpreauth

Remove the gak_fct, gak_data, salt, s2kparams, and as_key arguments
of krb5_clpreauth_process_fn and krb5_clpreauth_tryagain_fn.  To
replace them, add two callbacks: one which gets the AS key using the
previously selected etype-info2 information, and a second which lets
the module replace the AS key with one it has computed.

This changes limits module flexibility in a few ways.  Modules cannot
check whether the AS key was already obtained before asking for it,
and they cannot use the etype-info2 salt and s2kparams for purposes
other than getting the password-based AS key.  It is believed that
of existing preauth mechanisms, only SAM-2 preauth needs more
flexibility than the new interfaces provide, and as an internal legacy
mechanism it can cheat.  Future mechanisms should be okay since the
current IETF philosophy is that etype-info2 information should not be
used for other purposes.

ticket: 6976

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25351 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoDrop retransmits while processing requests
Greg Hudson [Sat, 15 Oct 2011 15:35:46 +0000 (15:35 +0000)]
Drop retransmits while processing requests

Supporting asynchronous preauth modules means that the KDC can receive
a retransmitted request before it finishes processing the initial
request.  Ignore those retransmits instead of processing them.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25350 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoUntabify kdc_preauth_encts.c
Greg Hudson [Sat, 15 Oct 2011 15:08:02 +0000 (15:08 +0000)]
Untabify kdc_preauth_encts.c

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25349 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoMake kdcpreauth edata method respond via callback
Greg Hudson [Sat, 15 Oct 2011 15:06:37 +0000 (15:06 +0000)]
Make kdcpreauth edata method respond via callback

From npmccallum@redhat.com with changes.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25348 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoMake get_preauth_hint_list respond via callback
Greg Hudson [Sat, 15 Oct 2011 15:03:17 +0000 (15:03 +0000)]
Make get_preauth_hint_list respond via callback

From npmccallum@redhat.com with changes.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25347 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoRemove enc-timestamp code from kdc_preauth.c
Greg Hudson [Sat, 15 Oct 2011 15:03:10 +0000 (15:03 +0000)]
Remove enc-timestamp code from kdc_preauth.c

This code should have been removed in r25319 but was not.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25346 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoExclude more stuff from make reindent
Tom Yu [Fri, 14 Oct 2011 23:14:53 +0000 (23:14 +0000)]
Exclude more stuff from make reindent

Apply exclusions to "make reindent" as well, to fully exclude some
files from whitespace cleanups.  Add fnmatch.c to exclusions.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25345 dc483132-0cff-0310-8789-dd5450dbe970

13 years agomake depend
Tom Yu [Fri, 14 Oct 2011 18:19:36 +0000 (18:19 +0000)]
make depend

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25344 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoUntabify a recent gssapi_alloc.h change
Greg Hudson [Fri, 14 Oct 2011 15:14:57 +0000 (15:14 +0000)]
Untabify a recent gssapi_alloc.h change

Also mark the file as using the krb5 C style.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25343 dc483132-0cff-0310-8789-dd5450dbe970

13 years agoFix gssapi_strdup
Sam Hartman [Fri, 14 Oct 2011 15:07:01 +0000 (15:07 +0000)]
Fix gssapi_strdup

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25342 dc483132-0cff-0310-8789-dd5450dbe970

13 years agogssalloc memory management for gss_buffer_set
Sam Hartman [Fri, 14 Oct 2011 14:46:57 +0000 (14:46 +0000)]
gssalloc memory management for gss_buffer_set

compiles, but untested

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25341 dc483132-0cff-0310-8789-dd5450dbe970

13 years agobuild profile dll (xpprof32/64.dll) on windows
Sam Hartman [Fri, 14 Oct 2011 14:44:35 +0000 (14:44 +0000)]
build profile dll (xpprof32/64.dll) on windows

Signed-off-by: Kevin Wasserman <kevin.wasserman@painless-security.com>
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@25340 dc483132-0cff-0310-8789-dd5450dbe970