Mon Oct 9 23:03:48 1995 Sam Hartman * kerberos5.c: make session_key a pointer, and use krb5_copy_keyblock not krb5_copy_keyblock_contents; there was no reason to violate this abstraction. Sun Sep 24 12:33:03 1995 Sam Hartman * kerberos5.c: Initialize session key from the subsession key we get from krb5_mk_req_extended, using ticket key as a fallback. (kerberos5_send): Use appropriate enctypes when encryption defined. Wed Sep 06 14:20:57 1995 Chris Provenzano (proven@mit.edu) * encrypt.h, kerberos5.c : s/keytype/enctype/g, s/KEYTYPE/ENCTYPE/g Tue Sep 05 22:10:34 1995 Chris Provenzano (proven@mit.edu) * kerberos5.c : Remove krb5_enctype references, and replace with krb5_keytype where appropriate. Thu Aug 3 11:36:15 EDT 1995 Paul Park (pjpark@mit.edu) * kerberos.c - Give the compiler something to compile when K4 disabled. Tue Jun 27 16:16:18 EDT 1995 Paul Park (pjpark@mit.edu) * enc_des.c, encrypt.c, krb4encpwd.c, read_password.c, rsaencpwd.c, spx.c - Give the compiler something to compile when these modules are essentially disabled. Some compilers choke when there's nothing to compile. * setenv.c - Change prototype for __findenv to be static since it's really static. Tue Jun 20 13:59:43 1995 Tom Yu (tlyu@dragons-lair) * configure.in: fix typo * strrchr.c: NO_STRING_H -> HAVE_STRING_H * strftime.c: NO_STRING_H -> HAVE_STRING_H * strerror.c: NO_STRING_H -> HAVE_STRING_H * strdup.c: NO_STRING_H -> HAVE_STRING_H * strchr.c: NO_STRING_H -> HAVE_STRING_H * strcasecmp.c: NO_STRING_H -> HAVE_STRING_H * spx.c: NO_STRING_H -> HAVE_STRING_H * rsaencpwd: NO_STRING_H -> HAVE_STRING_H * read_password.c: NO_STRING_H -> HAVE_STRING_H * mem.c: NO_STRING_H -> HAVE_STRING_H * krb4encpwd.c: NO_STRING_H -> HAVE_STRING_H * kerberos5.c: NO_STRING_H -> HAVE_STRING_H * kerberos.c: NO_STRING_H -> HAVE_STRING_H * encrypt.c: NO_STRING_H -> HAVE_STRING_H * auth.c: NO_STRING_H -> HAVE_STRING_H for consistency * configure.in: added missing tests for string.h, stdlib.h Sat Jun 10 22:59:42 1995 Tom Yu (tlyu@dragons-lair) * forward.c, kerberos5.c: krb5_auth_context redefinitions Fri Jun 9 18:30:02 1995 * configure.in: Remove standardized set of autoconf macros, which are now handled by CONFIG_RULES. Wed May 24 10:29:54 1995 Ezra Peisach * kerberos5.c: Include string.h/strings.h. Include stdlib.h or declare malloc. Sun May 7 18:45:09 1995 Ezra Peisach * kerberos5.c (kerberos5_send): Fix improperly closed comment krb5_get_credentials second argument is not kdc_options.... * configure.in (LIBOBJS): Removed duplicate WITH_KRB4 Fri Apr 28 11:17:16 1995 Mark Eichin * configure.in: switch to WITH_KRB4 since it suffices in this case. Thu Apr 27 17:08:16 1995 Mark Eichin * configure.in: use AC_CONST since we need it for v4. Thu Apr 27 15:52:19 1995 Chris Provenzano (proven@mit.edu) * kerberos5.c (kerberos_is()) : Initialize keytabid to NULL. Thu Apr 27 14:48:38 1995 Mark Eichin * Makefile.in (LOCALINCLUDES): find kerberosIV headers. Wed Apr 26 19:52:52 1995 Mark Eichin * kerberos5.c (kerberos5_is): use kt_resolve to get keytab, to correspond to current interface to rd_req. Tue Apr 25 21:23:28 1995 Chris Provenzano (proven@mit.edu) * forward.c (rd_and_store_for_creds()) : Rewritten to use auth_context and the new krb5_rd_creds(). * forward.c (get_for_creds()) : New function replacing krb5_get_for_creds() and uses auth_context and new krb5_mk_creds() routine. * kerberos5.c (kerberos5_send()): Set initial flags on auth_context to KRB5_AUTH_CONTEXT_RET_TIME, and use new rd_and_store_for_creds() routine. * kerberos5.c (kerberos5_forward()): Use the new get_for_creds(). Sat Apr 22 00:50:14 1995 Theodore Y. Ts'o (tytso@dcl) * kerberos5.c (kerberos5_init): Only call krb5_init_context if the telnet context hasn't been initialized yet. Thu Apr 20 20:12:32 1995 Mark Eichin Changes for testsuite from Ian Taylor * kerberos5.c (telnet_srvtab): New global variable. (telnet_krb5_realm): New global variable. (kerberos5_send): If telnet_krb5_realm is set, copy it into creds.server. Pass new_creds to krb5_mk_req_extended, not &creds. Pass &new_creds->keyblock to krb5_copy_keyblock_contents, not new_creds. (kerberos5_is): pass telnet_srvtab in to krb_rd_req. (kerberos5_forward): If telnet_krb5_realm is set, copy it into local_creds->server. Wed Mar 29 15:08:43 1995 Theodore Y. Ts'o (tytso@dcl) * kerberos5.c: No need to have the session_key established for mutual authentication to work. (That's only done if ENCRYPTION is defined.) * auth.c (authenticators): Allow mutual authentication even if the ENCRYPTION option is not turned on. Mon Mar 27 07:56:26 1995 Chris Provenzano (proven@mit.edu) * kerberos5.c (kerberos5_is()): Use new calling convention for krb5_rd_req(), and krb5_mk_rep(). Fri Mar 24 23:51:18 1995 Theodore Y. Ts'o * kerberos5.c (kerberos5_send): Initialize auth_context to zero before calling mk_req. Fri Mar 10 11:09:34 1995 Chris Provenzano (proven@mit.edu) * kerberos5.c: Use new calling convention for krb5_mk_req_extended(). Tue Mar 7 19:52:00 1995 Mark Eichin * configure.in: take out ISODE_DEFS, ISODE_INCLUDE. Tue Feb 28 01:48:32 1995 John Gilmore (gnu at toad.com) * forward.c, kerberos5.c: Avoid includes. Tue Feb 14 15:30:55 1995 Chris Provenzano (proven@mit.edu) * kerberos5.c (kerberos5_send(), kerberos5_forward()) Call krb5_get_credentials() and krb5_mk_req_extended() with new calling convention. Thu Feb 2 02:56:50 1995 John Gilmore * forward.c: Remove unused #include . * kerberos5.c (kerberos5_send): Remove code for sending a checksum of a zero-byte string; we can just send no checksum at all. This eliminates dependency on . (kerberos5_forward): Remove extra parameter to krb5_get_for_creds, probably accidentally inserted during context changes -- which don't seem to be here in the ChangeLog. * kerberos.c: Remove prototypes for krb4 functions, since some of them are wrong with CNS (u_long vs. KRB_INT32 conflicts). Fri Nov 18 15:19:26 1994 Theodore Y. Ts'o (tytso@dcl) * kerberos5.c (kerberos5_init): Initialize magic variable and encryption type. Fri Nov 18 00:37:13 1994 Mark Eichin * configure.in: use WITH_KRB4. (from epeisach) Mon Nov 14 16:27:29 1994 Theodore Y. Ts'o (tytso@dcl) * kerberos.c (kerberos4_is): Initialize random number generator on the server side so that the encryption routines later on can use it. * kerberos.c (kerberos4_send): Fix bug in how we pick the challenge for the challenge/response mutual authentication. Fri Nov 11 00:55:36 1994 Theodore Y. Ts'o (tytso@dcl) * forward.c (mk_cred, rd_cred): Move these routines to libkrb.a. Tue Nov 8 01:39:50 1994 Theodore Y. Ts'o (tytso@dcl) * kerberos.c (kerberos4_is): Fix bug in logic of incrementing the received challenge. A ++/-- mixup means there's a 1 in 256 chance the server will get it wrong. * kerberos.c: Use des_init_random_number_genator(), since that will result in different subsession keys on successive runs of telnet. Mon Nov 7 22:36:20 1994 Theodore Y. Ts'o (tytso@dcl) * auth.c (auth_status): Only print each possible authentication type once in the status report. * auth.c (auth_onoff): Remove excess call to getauthmask() which stomped the mask field. Only print each possible authentication type once in the help message. * auth.c (getauthmask): Fix reversed sense of strcasecmp comparison. * auth.c (auth_enable, auth_disable): Change the input type to be a char *, which is what auth_onoff wants anyway. Mon Aug 8 22:16:54 1994 Theodore Y. Ts'o (tytso at tsx-11) * kerberos5.c (kerberos5_send): Whoops, mispelled krb5_copy_keyblock_contents(). (It was inside #ifdef ENCRYPTION) Thu Aug 4 03:36:29 1994 Tom Yu (tlyu@dragons-lair) * Makefile.in: add blank target for install Tue Jul 26 18:21:29 1994 Tom Yu (tlyu@dragons-lair) * Makefile.in: whoops left out some $(srcdir) stuff Mon Jul 25 01:05:31 1994 Tom Yu (tlyu@dragons-lair) * Makefile.in: remove reference to lorder (linux doesn't have lorder, it seems) Fri Jul 15 23:36:50 1994 Theodore Y. Ts'o (tytso at tsx-11) * kerberos5.c (kerberos5_is): Avoid coredump caused by freeing of an unitialized variable. Also make sure we don't try to free name if it is NULL.