krb5.git
15 years agoCollect hairy parts of krb5_is_<msgtype> macros into one helper macro
Ken Raeburn [Mon, 2 Feb 2009 20:56:18 +0000 (20:56 +0000)]
Collect hairy parts of krb5_is_<msgtype> macros into one helper macro

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21869 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoRevise last change to better resemble the original test, keeping the
Ken Raeburn [Mon, 2 Feb 2009 20:37:41 +0000 (20:37 +0000)]
Revise last change to better resemble the original test, keeping the
min-lifetime test code collected together.  Change policy to have a
minimum password lifetime of 10s instead of 30s, and reduce the test
delays accordingly.

ticket: 6358

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21868 dc483132-0cff-0310-8789-dd5450dbe970

15 years agospeed up kpasswd tests
Ken Raeburn [Mon, 2 Feb 2009 19:29:52 +0000 (19:29 +0000)]
speed up kpasswd tests

Reorder some tests and tune delays, so that we don't need to run for
much more than twice the min-password-life interval when testing that
functionality.  (This could be made faster if we can assume that
init_db will always have been run immediately before the tests start.)

In my tests, this cuts something like 11 seconds off the run time (now
down to about 65 seconds).

ticket: 6358

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21867 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoaddress lib/kadm5 test suite slowness
Ken Raeburn [Mon, 2 Feb 2009 18:42:06 +0000 (18:42 +0000)]
address lib/kadm5 test suite slowness

In mod-principal tests for clearing the principal's policy, instead of
just testing to see if the wrong string is output and timing out
looking for it, check also for the new expected value.  Cuts test
suite run time by about two minutes for each pass (client vs server).

ticket: 6357
target_version: 1.7

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21866 dc483132-0cff-0310-8789-dd5450dbe970

15 years agodon't delay so long while waiting for daemon startup
Ken Raeburn [Mon, 2 Feb 2009 18:29:28 +0000 (18:29 +0000)]
don't delay so long while waiting for daemon startup

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21865 dc483132-0cff-0310-8789-dd5450dbe970

15 years agocomment needed some grammatical/typographical fixes too
Ken Raeburn [Mon, 2 Feb 2009 18:14:50 +0000 (18:14 +0000)]
comment needed some grammatical/typographical fixes too

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21864 dc483132-0cff-0310-8789-dd5450dbe970

15 years agocomment whitespace
Ken Raeburn [Mon, 2 Feb 2009 18:12:57 +0000 (18:12 +0000)]
comment whitespace

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21863 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoreapply memchr patch
Ken Raeburn [Mon, 2 Feb 2009 18:10:30 +0000 (18:10 +0000)]
reapply memchr patch

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21862 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoexplicitly initialize pointer fields after memset (for coverity)
Ken Raeburn [Mon, 2 Feb 2009 17:39:59 +0000 (17:39 +0000)]
explicitly initialize pointer fields after memset (for coverity)

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21861 dc483132-0cff-0310-8789-dd5450dbe970

15 years agosmall storage leak in KDC startup
Ken Raeburn [Mon, 2 Feb 2009 16:54:38 +0000 (16:54 +0000)]
small storage leak in KDC startup

Remove duplicate strdup call.

ticket: 6356
target_version: 1.7
tags: pullup

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21860 dc483132-0cff-0310-8789-dd5450dbe970

15 years agouse t_inetd with a ready message and avoid waiting a lot in non-root tests
Ken Raeburn [Sat, 31 Jan 2009 04:39:34 +0000 (04:39 +0000)]
use t_inetd with a ready message and avoid waiting a lot in non-root tests

Change t_inetd to print a ready message when it has started listening
on the indicated port number.

Look for this message in sample.exp rather than waiting an arbitrary
(and usually excessive) 2s each time for the inetd-mode tests.  Use
run_once to perform the standalone-mode test only once per test suite
invocation.

Change rsh and rcp tests to start the servers via t_inetd and avoid
excessive waiting at startup.

In some of my tests, this reduces the tests/dejagnu tests from taking
over 6 minutes to taking around 2 minutes.

(This does mean the server process will no longer have started up
before we launch the client, so it may be slower to respond, but it'll
still be faster than the 2s delay we used before even trying to
connect.)

We can probably eliminate the -D option code from krshd.c now.

The tests run as root (rlogin, telnet) still need updating.

ticket: 6355
target_version: 1.7
tags: pullup

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21855 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoback to trunk
Tom Yu [Sat, 31 Jan 2009 04:04:50 +0000 (04:04 +0000)]
back to trunk

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21854 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoREADME and patchlevel.h for 1.7 release branch
Tom Yu [Sat, 31 Jan 2009 04:00:10 +0000 (04:00 +0000)]
README and patchlevel.h for 1.7 release branch

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21852 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoDefault allow_weak_crypto=true for now. Default supported_enctypes to
Tom Yu [Sat, 31 Jan 2009 03:57:20 +0000 (03:57 +0000)]
Default allow_weak_crypto=true for now.  Default supported_enctypes to
exclude single-DES enctypes.

ticket: 6353
status: open

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21851 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoexport new functions
Ken Raeburn [Sat, 31 Jan 2009 02:55:25 +0000 (02:55 +0000)]
export new functions

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21850 dc483132-0cff-0310-8789-dd5450dbe970

15 years agofix resource leak
Ken Raeburn [Sat, 31 Jan 2009 01:50:37 +0000 (01:50 +0000)]
fix resource leak

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21849 dc483132-0cff-0310-8789-dd5450dbe970

15 years agorecode as more straight-line code to simplify analysis
Ken Raeburn [Sat, 31 Jan 2009 01:07:04 +0000 (01:07 +0000)]
recode as more straight-line code to simplify analysis

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21848 dc483132-0cff-0310-8789-dd5450dbe970

15 years agodon't use uninitialized variable
Ken Raeburn [Sat, 31 Jan 2009 00:54:33 +0000 (00:54 +0000)]
don't use uninitialized variable

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21847 dc483132-0cff-0310-8789-dd5450dbe970

15 years agohandle freeing null keyblock
Ken Raeburn [Sat, 31 Jan 2009 00:49:53 +0000 (00:49 +0000)]
handle freeing null keyblock

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21846 dc483132-0cff-0310-8789-dd5450dbe970

15 years agodon't double-free
Ken Raeburn [Sat, 31 Jan 2009 00:39:51 +0000 (00:39 +0000)]
don't double-free

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21845 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoMaster Key Migration Project
Will Fiveash [Fri, 30 Jan 2009 23:55:14 +0000 (23:55 +0000)]
Master Key Migration Project

Commit for the Master Key Migration Project.
http://k5wiki.kerberos.org/wiki/Projects/Master_Key_Migration

This commit provides the ability to add a new master key (with an
enctype differing from the current master key) to the master key
principal and stash file and then migrate the encryption of existing
principals long term keys to use the new master key.  In addition
deletion of master keys is provided.

ticket: 6354

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21844 dc483132-0cff-0310-8789-dd5450dbe970

15 years agokrb5_get_server_rcache had some dead (and incorrect) code for cleaning
Greg Hudson [Fri, 30 Jan 2009 21:25:40 +0000 (21:25 +0000)]
krb5_get_server_rcache had some dead (and incorrect) code for cleaning
up the rcache on failure.  Fix the cleanup code and make use of it.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21842 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoMake output parameter value of krb5_rc_resolve_full well-defined on
Greg Hudson [Fri, 30 Jan 2009 21:22:31 +0000 (21:22 +0000)]
Make output parameter value of krb5_rc_resolve_full well-defined on
error return.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21841 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIn krb5_kdcrep2creds, call the correct free function when cleaning up
Greg Hudson [Fri, 30 Jan 2009 21:21:10 +0000 (21:21 +0000)]
In krb5_kdcrep2creds, call the correct free function when cleaning up
the keyblock, and clean up the keyblock if krb5_copy_data fails.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21840 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoChanged the name match_config_pattern to krb5_match_config_pattern. Check strdup...
Zhanna Tsitkov [Fri, 30 Jan 2009 20:40:46 +0000 (20:40 +0000)]
Changed the name match_config_pattern to krb5_match_config_pattern. Check strdup return code.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21838 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIntroduced match_config_pattern function to find a pattern in the config value strings
Zhanna Tsitkov [Thu, 29 Jan 2009 19:23:22 +0000 (19:23 +0000)]
Introduced match_config_pattern function to find a pattern in the config value strings

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21828 dc483132-0cff-0310-8789-dd5450dbe970

15 years agowhitespace
Ken Raeburn [Thu, 29 Jan 2009 01:19:01 +0000 (01:19 +0000)]
whitespace

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21826 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoclear outbuf[0].data to unconfuse coverity
Ken Raeburn [Thu, 29 Jan 2009 01:12:17 +0000 (01:12 +0000)]
clear outbuf[0].data to unconfuse coverity

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21825 dc483132-0cff-0310-8789-dd5450dbe970

15 years agodisable single-DES by default
Tom Yu [Wed, 28 Jan 2009 23:22:27 +0000 (23:22 +0000)]
disable single-DES by default

Mark all single-DES enctypes as "weak", and create a new libdefaults
variable "allow_weak_crypto", which defaults to "false".

ticket: 6353
status: open

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21823 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoinclude netinet/in.h to make sure we know what struct in_addr looks like
Ken Raeburn [Wed, 28 Jan 2009 17:01:36 +0000 (17:01 +0000)]
include netinet/in.h to make sure we know what struct in_addr looks like

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21820 dc483132-0cff-0310-8789-dd5450dbe970

15 years agono more -r option for setting krb.conf path
Ken Raeburn [Wed, 28 Jan 2009 05:42:11 +0000 (05:42 +0000)]
no more -r option for setting krb.conf path

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21819 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoremove some remnants of krb4-related config file options
Ken Raeburn [Wed, 28 Jan 2009 05:37:10 +0000 (05:37 +0000)]
remove some remnants of krb4-related config file options

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21818 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoreturn correct kvno in TGS case
Sam Hartman [Wed, 28 Jan 2009 01:28:25 +0000 (01:28 +0000)]
return correct kvno in TGS case

krb5_c_encrypt sets the kvno of ciphertext to 0.
So fill it in after the call to encrypt_tkt_part.

ticket: 6352

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21815 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoRun a single expect statement to read from both client and server, to
Tom Yu [Wed, 28 Jan 2009 00:32:29 +0000 (00:32 +0000)]
Run a single expect statement to read from both client and server, to
avoid an apparent race condition on Darwin.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21814 dc483132-0cff-0310-8789-dd5450dbe970

15 years agowhitespace
Ken Raeburn [Wed, 28 Jan 2009 00:30:06 +0000 (00:30 +0000)]
whitespace

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21813 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoStart to phase out krb5_xfree macro, which just casts its argument to
Ken Raeburn [Wed, 28 Jan 2009 00:04:15 +0000 (00:04 +0000)]
Start to phase out krb5_xfree macro, which just casts its argument to
char* and calls free.

Replace most uses, outside of the LDAP KDB plugin, which doesn't build
on my test system of the moment because of version dependencies.  Add
one explicit cast to make the change warning-neutral (under gcc 4.0.1
on Mac OS X 10.5.6).

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21812 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoHelper script for processing valgrind logs that don't indicate any errors, or are...
Ken Raeburn [Tue, 27 Jan 2009 23:14:35 +0000 (23:14 +0000)]
Helper script for processing valgrind logs that don't indicate any errors, or are for system programs

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21810 dc483132-0cff-0310-8789-dd5450dbe970

15 years agouse memchr
Ken Raeburn [Tue, 27 Jan 2009 22:00:15 +0000 (22:00 +0000)]
use memchr

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21809 dc483132-0cff-0310-8789-dd5450dbe970

15 years ago"Coding practices" related fixes
Zhanna Tsitkov [Tue, 27 Jan 2009 21:09:35 +0000 (21:09 +0000)]
"Coding practices" related fixes

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21808 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIn asn1_decode_pa_for_user(), ensure that user member is allocated
Luke Howard [Mon, 26 Jan 2009 22:50:02 +0000 (22:50 +0000)]
In asn1_decode_pa_for_user(), ensure that user member is allocated
before assigning a value to it.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21805 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIntroduced new static function prep_reprocess_req to make code more readable
Zhanna Tsitkov [Mon, 26 Jan 2009 21:26:33 +0000 (21:26 +0000)]
Introduced new static function prep_reprocess_req to make code more readable

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21801 dc483132-0cff-0310-8789-dd5450dbe970

15 years agogss_header|trailerlen should be unsigned int
Sam Hartman [Mon, 26 Jan 2009 20:02:54 +0000 (20:02 +0000)]
gss_header|trailerlen should be unsigned int

The krb5_c_crypto_length API returns unsigned int per its design.
so, use unsigned int not size_t for its output.

ticket: 6351

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21799 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPreliminary update of README for krb5-1.7 release
Tom Yu [Mon, 26 Jan 2009 19:47:57 +0000 (19:47 +0000)]
Preliminary update of README for krb5-1.7 release

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21798 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoUse 16/32-bit big/little-endian store functions in more places
Ken Raeburn [Mon, 26 Jan 2009 19:06:21 +0000 (19:06 +0000)]
Use 16/32-bit big/little-endian store functions in more places

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21796 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoCheck asprintf return codes. Styling
Zhanna Tsitkov [Mon, 26 Jan 2009 16:20:41 +0000 (16:20 +0000)]
Check asprintf return codes. Styling

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21795 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoUse a struct in_addr to insure alignment of address - instead of
Ezra Peisach [Sun, 25 Jan 2009 16:44:02 +0000 (16:44 +0000)]
Use a struct in_addr to insure alignment of address - instead of
random alignment on the stack. Solaris 2.10 has issues if the address
is not aligned.  The rest of the code in the tree uses a struct
in_addr or mallocs the address - which will be sufficiently aligned.

ticket: 6308

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21794 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoChange 16/32/64-bit big-/little-endian/native unaligned load/store
Ken Raeburn [Sat, 24 Jan 2009 01:27:16 +0000 (01:27 +0000)]
Change 16/32/64-bit big-/little-endian/native unaligned load/store
routines to take void pointers, so they can operate on both plain and
unsigned char buffers, or other types.
Remove some now-unneeded casts.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21793 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoEliminate the need for the domain_realm mapping table on the client side by implement...
Zhanna Tsitkov [Fri, 23 Jan 2009 19:59:50 +0000 (19:59 +0000)]
Eliminate the need for the domain_realm mapping table on the client side by implementing minimal referral support in the KDC

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21792 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoRemove some null checks in cases where pointers can't be null
Greg Hudson [Fri, 23 Jan 2009 19:04:57 +0000 (19:04 +0000)]
Remove some null checks in cases where pointers can't be null

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21790 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIn krb5_pac_get_buffer, check the correct value for null after calling
Greg Hudson [Fri, 23 Jan 2009 19:01:56 +0000 (19:01 +0000)]
In krb5_pac_get_buffer, check the correct value for null after calling
malloc.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21789 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIn krb5_pac_parse, change the code around a bit to avoid harmlessly
Greg Hudson [Fri, 23 Jan 2009 18:41:39 +0000 (18:41 +0000)]
In krb5_pac_parse, change the code around a bit to avoid harmlessly
copying an uninitialized Buffers field of a PACTYPE structure.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21788 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoInitialize ds.magic in k5_md5_hmac_hash, to avoid harmlessly copying
Greg Hudson [Fri, 23 Jan 2009 18:19:19 +0000 (18:19 +0000)]
Initialize ds.magic in k5_md5_hmac_hash, to avoid harmlessly copying
around its uninitialized value in krb5_hmac.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21787 dc483132-0cff-0310-8789-dd5450dbe970

15 years agohash db2 code breaks if st_blksize > 64k
Ken Raeburn [Fri, 23 Jan 2009 18:04:08 +0000 (18:04 +0000)]
hash db2 code breaks if st_blksize > 64k

The hash db code assumes in places that the block size is no larger
than 64K.  There's a range check in the case where you don't have a
file but provide initialization info.  The btree code will cap the
block size used at 64K.

Apparently Sun's ZFS can report back a block size of 128K, causing the
db2 tests to fail.

Add such a cap to the hash db creation code.

Note that our default configuration is to use the btree code when
creating a new database, so it's unlikely that this will cause
real-world problems unless someone went out of their way to specify
use of the hash format.

ticket: 6342

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21786 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIn krb5_get_init_creds, null out err_reply in a case where we free it
Greg Hudson [Fri, 23 Jan 2009 17:53:52 +0000 (17:53 +0000)]
In krb5_get_init_creds, null out err_reply in a case where we free it
and do not necessarily exit the loop.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21785 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoAlways initialize the output parameter in krb5_parse_name and friends
Greg Hudson [Fri, 23 Jan 2009 17:45:58 +0000 (17:45 +0000)]
Always initialize the output parameter in krb5_parse_name and friends

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21784 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIn kadmin, free the correct item in an error case (fixes a double-free
Greg Hudson [Fri, 23 Jan 2009 17:24:50 +0000 (17:24 +0000)]
In kadmin, free the correct item in an error case (fixes a double-free
and a memory leak).

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21783 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoAdd some output parameter initializations n order to eliminate some
Greg Hudson [Fri, 23 Jan 2009 06:51:03 +0000 (06:51 +0000)]
Add some output parameter initializations n order to eliminate some
spurious Coverity defects.  (Far from a comprehensive pass.)

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21782 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPer coding standards, don't declare variables in inner scopes in new
Greg Hudson [Fri, 23 Jan 2009 05:02:07 +0000 (05:02 +0000)]
Per coding standards, don't declare variables in inner scopes in new
rcache code except when it really makes sense (option processing for
t_replay).

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21781 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoInclude cksumtypes.h from aead.h to get struct krb5_cksumtypes (moved
Ken Raeburn [Fri, 23 Jan 2009 00:46:46 +0000 (00:46 +0000)]
Include cksumtypes.h from aead.h to get struct krb5_cksumtypes (moved
in rev 21753).
Protect cksumtypes.h from multiple inclusions.
Update dependencies.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21780 dc483132-0cff-0310-8789-dd5450dbe970

15 years agofix trailing whitespace
Tom Yu [Thu, 22 Jan 2009 23:37:35 +0000 (23:37 +0000)]
fix trailing whitespace

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21779 dc483132-0cff-0310-8789-dd5450dbe970

15 years agowhitespace
Tom Yu [Thu, 22 Jan 2009 23:21:11 +0000 (23:21 +0000)]
whitespace

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21778 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoAdapted patch from Apple: in kadmind's process_chpw_request, make sure
Greg Hudson [Thu, 22 Jan 2009 19:19:34 +0000 (19:19 +0000)]
Adapted patch from Apple: in kadmind's process_chpw_request, make sure
to free error message strings.

ticket: 6284
status: open

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21776 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIf USE_VALGRIND, check that source principal is defined before copying.
Ken Raeburn [Thu, 22 Jan 2009 06:34:14 +0000 (06:34 +0000)]
If USE_VALGRIND, check that source principal is defined before copying.
Initialize magic number fields of allocated name components.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21775 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIf USE_VALGRIND, check readability of byte string before copying
Ken Raeburn [Thu, 22 Jan 2009 06:30:08 +0000 (06:30 +0000)]
If USE_VALGRIND, check readability of byte string before copying

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21774 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoDefine valgrind macros as no-ops if not USE_VALGRIND, and use unconditionally, per...
Ken Raeburn [Thu, 22 Jan 2009 06:04:10 +0000 (06:04 +0000)]
Define valgrind macros as no-ops if not USE_VALGRIND, and use unconditionally, per Danilo's suggestion

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21773 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPatch from Love: when opening an existing replay cache, check more
Greg Hudson [Wed, 21 Jan 2009 19:38:12 +0000 (19:38 +0000)]
Patch from Love: when opening an existing replay cache, check more
thoroughly to prevent symlink attacks.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21770 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPatch from Luke: fix error return of krb5_gss_use_kdc_context
Greg Hudson [Wed, 21 Jan 2009 18:23:58 +0000 (18:23 +0000)]
Patch from Luke: fix error return of krb5_gss_use_kdc_context

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21768 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoWhen encoding, and compiling with -DUSE_VALGRIND, explicitly get valgrind to check...
Ken Raeburn [Tue, 20 Jan 2009 23:49:08 +0000 (23:49 +0000)]
When encoding, and compiling with -DUSE_VALGRIND, explicitly get valgrind to check the inputs

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21766 dc483132-0cff-0310-8789-dd5450dbe970

15 years agodelete trailing whitespace, including insanely long 'blank' lines
Ken Raeburn [Tue, 20 Jan 2009 22:52:15 +0000 (22:52 +0000)]
delete trailing whitespace, including insanely long 'blank' lines

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21764 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIn krb5_rd_rep_dce, initialize repl to NULL to avoid using its
Greg Hudson [Tue, 20 Jan 2009 18:50:51 +0000 (18:50 +0000)]
In krb5_rd_rep_dce, initialize repl to NULL to avoid using its
initialized value when we take the first goto to clean_scratch.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21763 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIf buffer length is 0, force pointer to NULL, to keep valgrind happier when callers...
Ken Raeburn [Sat, 17 Jan 2009 01:00:22 +0000 (01:00 +0000)]
If buffer length is 0, force pointer to NULL, to keep valgrind happier when callers don't bother

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21762 dc483132-0cff-0310-8789-dd5450dbe970

15 years agopatch from Luke - minor cleanup
Ken Raeburn [Sat, 17 Jan 2009 00:06:06 +0000 (00:06 +0000)]
patch from Luke - minor cleanup

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21761 dc483132-0cff-0310-8789-dd5450dbe970

15 years agopatch from Luke - fix enctype-nego enctype list setup
Ken Raeburn [Sat, 17 Jan 2009 00:04:49 +0000 (00:04 +0000)]
patch from Luke - fix enctype-nego enctype list setup

ticket: 6336

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21760 dc483132-0cff-0310-8789-dd5450dbe970

15 years agopatch from Luke - fix sign problem
Ken Raeburn [Sat, 17 Jan 2009 00:02:05 +0000 (00:02 +0000)]
patch from Luke - fix sign problem

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21759 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoinclude dependencies for decoder test
Ken Raeburn [Fri, 16 Jan 2009 23:55:01 +0000 (23:55 +0000)]
include dependencies for decoder test

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21758 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoUse valgrind in more cases if VALGRIND is set
Ken Raeburn [Fri, 16 Jan 2009 00:33:00 +0000 (00:33 +0000)]
Use valgrind in more cases if VALGRIND is set

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21754 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoMake enctype and checksumtype name mapping table types private to the
Ken Raeburn [Thu, 15 Jan 2009 23:22:53 +0000 (23:22 +0000)]
Make enctype and checksumtype name mapping table types private to the
crypto library.  Add a field for aliases to the record type so that
aliases don't have to be implemented by duplicating all the other
information.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21753 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoFix an additional multiple-write case noted by John, where sendauth
Ken Raeburn [Thu, 15 Jan 2009 19:15:22 +0000 (19:15 +0000)]
Fix an additional multiple-write case noted by John, where sendauth
calls write_message twice in a row.

Add new function krb5int_write_messages, calls krb5_net_writev with
multiple messages (currently only two at a time).  Use it from
krb5_write_message and krb5_sendauth.

ticket: 6339

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21752 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoRework the replay cache extensions to make the hash extension records
Greg Hudson [Thu, 15 Jan 2009 19:11:45 +0000 (19:11 +0000)]
Rework the replay cache extensions to make the hash extension records
stand alone.  Otherwise, reordering of records during an expunge could
cause the hash to be applied to the wrong record.

Also add an "expunge" option to the t_replay program, and clean up some
memory-handling inconsistencies.

ticket: 1201

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21751 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoupdate
Ken Raeburn [Thu, 15 Jan 2009 01:13:04 +0000 (01:13 +0000)]
update

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21750 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoAdd new routine krb5int_net_writev using scatter-gather source.
Ken Raeburn [Thu, 15 Jan 2009 00:59:27 +0000 (00:59 +0000)]
Add new routine krb5int_net_writev using scatter-gather source.
Use it from krb5_net_write to ensure testing and reduce duplication.
Use it from krb5_write_message to avoid Nagle+DelayedAck problem.

ticket: 6339

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21749 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIf we have a local UDP socket without the PKTINFO option set, it's
Ken Raeburn [Wed, 14 Jan 2009 23:36:04 +0000 (23:36 +0000)]
If we have a local UDP socket without the PKTINFO option set, it's
bound to a local address, so use getsockname to extract the local
(destination) address.

ticket: 6335

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21748 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoIf recv_from_to is passed a buffer for the local endpoint address,
Ken Raeburn [Wed, 14 Jan 2009 21:29:10 +0000 (21:29 +0000)]
If recv_from_to is passed a buffer for the local endpoint address,
clobber it before doing anything else, just in case we can't retrieve
the address and the caller blindly uses the buffer anyways.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21747 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoMerge some very simple points of divergence in the two copies of network.c -- enum...
Ken Raeburn [Wed, 14 Jan 2009 20:05:09 +0000 (20:05 +0000)]
Merge some very simple points of divergence in the two copies of network.c -- enum ordering, whitespace,
duplicate macro definitions, unused code, 0 vs NULL...

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21746 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPass s4u name and c_flags to log_tgs_req. If values are supplied, log
Ken Raeburn [Wed, 14 Jan 2009 19:50:34 +0000 (19:50 +0000)]
Pass s4u name and c_flags to log_tgs_req.  If values are supplied, log
an additional message to record the name and s4u mode.

Untested for lack of code to invoke these code paths.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21745 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoForce tickets acquired by the kadm5 client library via password
Russ Allbery [Wed, 14 Jan 2009 00:29:04 +0000 (00:29 +0000)]
Force tickets acquired by the kadm5 client library via password
authentication to be non-forwardable and non-proxiable, overridding
any [libdefaults] configuration.  This may be necessary at sites that
set forwardable to true by default in their krb5.conf files but
disable forwardable tickets for privileged principals.  Since the
ticket cache acquired by the kadm5 client library is used only for
kadmin operations, where forwardable is not useful or necessary, there
is no reason to ever attempt to obtain forwardable or proxiable tickets
here.

Ticket: 6337

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21744 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPatch from Luke Howard to:
Sam Hartman [Tue, 13 Jan 2009 22:57:42 +0000 (22:57 +0000)]
Patch from Luke Howard  to:
* Accept both CFX and non-CFX tokens all the time on acceptor
* Only produce an acceptor subkey if you are using cfx or dce or negotiating up to cfx

Additional changes from Sam Hartman:
* do not assume that the ticket  key type (server key) is a valid target for negotiation: the client may not support it.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21742 dc483132-0cff-0310-8789-dd5450dbe970

15 years ago/tmp/3
Ken Raeburn [Tue, 13 Jan 2009 21:54:45 +0000 (21:54 +0000)]
/tmp/3

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21741 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoFreeBSD compiler errors out on an error "zero or negative size array"
Ezra Peisach [Tue, 13 Jan 2009 19:43:18 +0000 (19:43 +0000)]
FreeBSD compiler errors out on an error "zero or negative size array"
after setting up an array with no elements.

ifdef out array declarations and code that uses it until there are entries.

Affects: krb5_gss_inquire_cred_by_oid_ops and
krb5_gss_set_sec_context_option_ops which would return an error in any
case as here are no entries in the arrays.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21740 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoAdd a call to limit_string that appears to have been accidentally removed
Sam Hartman [Mon, 12 Jan 2009 21:18:45 +0000 (21:18 +0000)]
Add a call to limit_string that appears to have been accidentally removed
somewhere along the mskrb-integ branch

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21729 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoRemove gss_export_name_object and gss_import_name_object.
Sam Hartman [Mon, 12 Jan 2009 21:03:02 +0000 (21:03 +0000)]
Remove gss_export_name_object and gss_import_name_object.
These are not standard interfaces, are not used by our tree
and were added because they might be useful but ended up not being used.
The stubs in gssapi.hin remain as they were shipped with previous releases.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21728 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPatch from Luke Howard:
Sam Hartman [Mon, 12 Jan 2009 19:59:16 +0000 (19:59 +0000)]
Patch from Luke Howard:
Previously when using the kdb keytab, there was a check to confirm that the server
was supported as a server and that attackers
could not force an enctype downgrade.
Add these to kdc_get_server_key

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21727 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoPatch from Luke Howard
Sam Hartman [Mon, 12 Jan 2009 19:43:13 +0000 (19:43 +0000)]
Patch from Luke Howard
There's a superfluous check in kdc_util.c; ad_entry is always non NULL

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21726 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoRestore behavior of returning KRB5APP_ERR_BAD_INTEGRITY from
Sam Hartman [Mon, 12 Jan 2009 19:43:07 +0000 (19:43 +0000)]
Restore behavior of returning KRB5APP_ERR_BAD_INTEGRITY from
preauth methods.
This creates a problem for Windows clients, but not doing it creates a problem for MIT clients.
Today our KDC is more likely to be used with MIT clients, but we need to examine this issues in more detail.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21725 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoCheck the return code from krb5int_clean_hostname as the sanity verification of...
Zhanna Tsitkov [Mon, 12 Jan 2009 19:40:37 +0000 (19:40 +0000)]
Check the return code from krb5int_clean_hostname as the  sanity verification of the hostname might fail

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21724 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoAdd message hash support to the replay interface, using extension
Greg Hudson [Mon, 12 Jan 2009 18:29:42 +0000 (18:29 +0000)]
Add message hash support to the replay interface, using extension
records (with an empty client string) to retain compatibility with old
code.  For rd_req, the ciphertext of the authenticator (with no ASN.1
wrapping) is hashed; for other uses of the replay cache, no message
hash is used at this time.

This commit adds a command-line tool for testing the replay cache but
does not add any automated tests.

ticket: 1201

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21723 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoFollow "off-path" TGT referrals
Tom Yu [Fri, 9 Jan 2009 01:42:37 +0000 (01:42 +0000)]
Follow "off-path" TGT referrals

ticket: 5627
status: open

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21720 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoRemove conflict marker; restore broken copyright line
Ken Raeburn [Fri, 9 Jan 2009 00:07:50 +0000 (00:07 +0000)]
Remove conflict marker; restore broken copyright line

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21719 dc483132-0cff-0310-8789-dd5450dbe970

15 years agokdb/keytab.c: map KRB5_KDB_NO_MATCHING_KEY to KRB5_KT_KVNONOTFOUND.
Sam Hartman [Wed, 7 Jan 2009 18:13:30 +0000 (18:13 +0000)]
kdb/keytab.c: map KRB5_KDB_NO_MATCHING_KEY to KRB5_KT_KVNONOTFOUND.
At least in cases other than tgts, this  code handles its own enctype matching, so kvno not found is the only
thing that produces the no matching key error.

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21718 dc483132-0cff-0310-8789-dd5450dbe970

15 years agoDon't create include/kerberosIV on installation
Ken Raeburn [Wed, 7 Jan 2009 00:11:40 +0000 (00:11 +0000)]
Don't create include/kerberosIV on installation

git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@21717 dc483132-0cff-0310-8789-dd5450dbe970