On 07/17/2017 04:59 PM, Greg Hudson wrote:
> (Sent unicast.)
>
> Hm, our mailing list software seems to have removed all of the content
> from both of your messages, due to some incompatibility with the way
> they were formatted. Would it be possible to combine them and resend
> them as plain text? Unfortunately I no longer have a copy of the
> contents after moderating them through.
>
Sure, no problem, here they are. Let me know if there are still issues with
getting my content. I sent this one in plaintext:
I ran the kdb5_util program under valgrind as well and saw this, thought I'd
pass it along:
root@bllkrb501:~# valgrind kdb5_util stash -f /etc/krb5kdc/stash
==16389== Memcheck, a memory error detector
==16389== Copyright (C) 2002-2015, and GNU GPL'd, by Julian Seward et al.
==16389== Using Valgrind-3.11.0 and LibVEX; rerun with -h for copyright info
==16389== Command: kdb5_util stash -f /etc/krb5kdc/stash
==16389==
stat(/usr/local/lib/krb5/plugins/kdb/kldap): No such file or directory
get_plugin_data_sym(kdb_function_table)
==16389== Warning: invalid file descriptor -1 in syscall write()
kdb5_util: Cannot find master key record in database while getting master
key list
kdb5_util: Warning: proceeding without master key list
Enter KDC database master key:
==16389== Invalid read of size 2
==16389== at 0x506DFA8: krb5_db_fetch_mkey (kdb5.c:1224)
==16389== by 0x406D56: kdb5_stash (kdb5_stash.c:110)
==16389== by 0x4048F1: main (kdb5_util.c:346)
==16389== Address 0x2 is not stack'd, malloc'd or (recently) free'd
==16389==
==16389==
==16389== Process terminating with default action of signal 11 (SIGSEGV)
[...]
NULL pointer?
On 07/17/2017 03:35 PM, Joshua Schaeffer wrote:
> TL;DR
> I'm getting a segmentation error when I run kdb5_util stash from a compiled
> version of 1.15.1:
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x00007ffff799afa8 in krb5_db_fetch_mkey (context=0x61eb80,
> mname=0x678a60, etype=18, fromkeyboard=1, twice=0,
> db_args=0x0, kvno=0x7fffffffe56c, salt=0x0, key=0x619c30
> <master_keyblock>) at kdb5.c:1224
> 1224 *kvno = (krb5_kvno)
> master_entry->key_data->key_data_kvno;
>
> --------------------------------------------------------
>
> Hey all,
>
> I'm trying to figure out why I'm getting a segmentation fault when I try to
> start the krb5-admin-server service. I have a server running in an LXD
> container, which I think is causing the issue, but I'm not sure what the
> container doesn't have permissions/rights to that is causing this problem and
> I've searched all my log files far and wide and can't find any smoking gun.
> So here is what I've done:
>
> First I tried installing MIT Kerberos using the package management system
> which installs version 1.13.2. Then I setup my krb5.conf file and initialize
> my database:
>
> kdb5_ldap_util -D cn=admin,dc=appendata,dc=net create -subtrees 'ou=End
> Users,ou=People,dc=appendata,dc=net':'ou=Other
> Users,ou=People,dc=appendata,dc=net -r APPENDATA.NET -s -H
> ldaps://bllldap01.appendata.net
>
> This works without issue, so I proceed by stashing a few ldap user's
> passwords, create my kadm5.acl file and then I go and try to start kadmind:
>
> root@bllkrb501:~# systemctl start krb5-admin-server
> root@bllkrb501:~# systemctl status krb5-admin-server
> krb5-admin-server.service - Kerberos 5 Admin Server
> Loaded: loaded (/lib/systemd/system/krb5-admin-server.service;
> enabled; vendor preset: enabled)
> Drop-In: /lib/systemd/system/krb5-admin-server.service.d
> └─slapd-before-kdc.conf
> Active: failed (Result: core-dump) since Mon 2017-07-17 15:00:36 MDT;
> 6s ago
> Process: 3304 ExecStart=/usr/sbin/kadmind -nofork $DAEMON_ARGS
> (code=dumped, signal=SEGV)
> Main PID: 3304 (code=dumped, signal=SEGV)
>
> Jul 17 15:00:35 bllkrb501 systemd[1]: Started Kerberos 5 Admin Server.
> Jul 17 15:00:36 bllkrb501 systemd[1]: krb5-admin-server.service: Main
> process exited, code=dumped, status=11/SEGV
> Jul 17 15:00:36 bllkrb501 systemd[1]: krb5-admin-server.service: Unit
> entered failed state.
> Jul 17 15:00:36 bllkrb501 systemd[1]: krb5-admin-server.service: Failed
> with result 'core-dump'.
>
> And if I try to start kadmind manually:
>
> root@bllkrb501:~# kadmind -nofork
> Segmentation fault (core dumped)
>
> Here is an strace of the same command:
>
> [...]
> write(11,
> "\27\3\3\2Y\0\0\0\0\0\0\0\3ZMi\3049\2256\337\17y}\361\237\4Kv\f\347\233"...,
> 606) = 606
> poll([{fd=11, events=POLLIN|POLLPRI}], 1, 300000) = 1 ([{fd=11,
> revents=POLLIN}])
> read(11, "\27\3\3\0&", 5) = 5
> read(11,
> "\0\0\0\0\0\0\0\4\313(H\177\362\376\4\34\251\266T\23\5\ndj\327\311\304\30\177\31\26b"...,
> 38) = 38
> write(11,
> "\27\3\3\2[\0\0\0\0\0\0\0\4\244G3\341}F\35:\340\244\356\250\254T\365g\7\240r"...,
> 608) = 608
> poll([{fd=11, events=POLLIN|POLLPRI}], 1, 300000) = 1 ([{fd=11,
> revents=POLLIN}])
> read(11, "\27\3\3\0&", 5) = 5
> read(11,
> "\0\0\0\0\0\0\0\5\4\204S\v9\305v\217\324\r\316\313\207\2405\245\2749\242\356\341\361h\367"...,
> 38) = 38
> write(11,
> "\27\3\3\2g\0\0\0\0\0\0\0\5\34\306\243F\177zh\370s\352\230\206\243\215\345\3719\\_"...,
> 620) = 620
> poll([{fd=11, events=POLLIN|POLLPRI}], 1, 300000) = 1 ([{fd=11,
> revents=POLLIN}])
> read(11, "\27\3\3\2\27", 5) = 5
> read(11,
> "\0\0\0\0\0\0\0\6\f\332:\226l\34J\0\344v\304K\203\242\0\356[X~\225\347\253\37P"...,
> 535) = 535
> poll([{fd=11, events=POLLIN|POLLPRI}], 1, 299999) = 1 ([{fd=11,
> revents=POLLIN}])
> read(11, "\27\3\3\0&", 5) = 5
> read(11,
> "\0\0\0\0\0\0\0\7v\215\202\33\312\325\316xL4&\305i^\310\21,X\226\211\357\317\323\354"...,
> 38) = 38
> open("/etc/localtime", O_RDONLY|O_CLOEXEC) = 23
> fstat(23, {st_mode=S_IFREG|0644, st_size=2453, ...}) = 0
> fstat(23, {st_mode=S_IFREG|0644, st_size=2453, ...}) = 0
> read(23,
> "TZif2\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\5\0\0\0\5\0\0\0\0"..., 2560) = 2453
> lseek(23, -1559, SEEK_CUR) = 894
> read(23,
> "TZif2\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\5\0\0\0\5\0\0\0\0"..., 2560) = 1559
> close(23) = 0
> --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x4} ---
> +++ killed by SIGSEGV (core dumped) +++
> Segmentation fault (core dumped)
>
> I wasn't able to determine too much from this. To me it looks like the system
> is opening /etc/localtime and then the program crashes. Next I resorted to
> debugging the error and that is where I'm currently at. I downloaded the
> source code for 1.15.1, installed it, and reran through all my steps. I
> wasn't able to get to my step of trying to start kadmind, because it
> complains the stash file doesn't exist for the master key, which it doesn't,
> and I'm not sure why it isn't created when I issued my krb5_ldap_util command
> above, but I was able to still get a segmentation fault when I try to create
> the stash file:
>
> root@bllkrb501:~# kdb5_util stash
> stat(/usr/local/lib/krb5/plugins/kdb/kldap): No such file or directory
> get_plugin_data_sym(kdb_function_table)
> kdb5_util: Cannot find master key record in database while getting master
> key list
> kdb5_util: Warning: proceeding without master key list
> Enter KDC database master key:
> Segmentation fault (core dumped)
>
> root@bllkrb501:~# ls -l /usr/local/lib/krb5/plugins/kdb/
> total 407
> -rw-r--r-- 1 root root 366680 Jul 17 12:51 db2.so
> -rw-r--r-- 1 root root 21008 Jul 17 12:51 kldap.so
>
> I'm not sure why it is complaining about plugins/kdb/kldap not existing
> either. The shared object exists under that directory. Perhaps this is the
> problem. I compiled Kerberos with "--with-ldap". I've also run the same
> command through gdb and got the line it is failing at:
>
> root@bllkrb501:~# gdb kdb5_util
> [...]
> Reading symbols from kdb5_util...done.
> (gdb) run stash -f /etc/krb5kdc/stash
> Starting program: /usr/local/sbin/kdb5_util stash -f /etc/krb5kdc/stash
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
> stat(/usr/local/lib/krb5/plugins/kdb/kldap): No such file or directory
> get_plugin_data_sym(kdb_function_table)
> kdb5_util: Cannot find master key record in database while getting master
> key list
> kdb5_util: Warning: proceeding without master key list
> Enter KDC database master key:
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x00007ffff799afa8 in krb5_db_fetch_mkey (context=0x61eb80,
> mname=0x678a60, etype=18, fromkeyboard=1, twice=0,
> db_args=0x0, kvno=0x7fffffffe56c, salt=0x0, key=0x619c30
> <master_keyblock>) at kdb5.c:1224
> 1224 *kvno = (krb5_kvno)
> master_entry->key_data->key_data_kvno;
> (gdb) continue
> Continuing.
>
> Program terminated with signal SIGSEGV, Segmentation fault.
> The program no longer exists.
> (gdb) quit
>
> I looked at the code and this is where it is actually failing:
>
> 1218 if (kvno != NULL && *kvno == IGNORE_VNO) {
> 1219 krb5_error_code rc;
> 1220 krb5_db_entry *master_entry;
>
> 1222 rc = krb5_db_get_principal(context, mname, 0,
> &master_entry);
> 1223 if (rc == 0) {
> 1224 *kvno = (krb5_kvno)
> master_entry->key_data->key_data_kvno;
> 1225 krb5_db_free_principal(context, master_entry);
> 1226 } else
> 1227 *kvno = 1;
> 1228 }
>
> I don't really know where to go from here. I don't know this code well enough
> to figure out why the segmentation error is occurring. Can anybody help me
> out? If you need additional information, I'd be happy to provide.
>
> Thanks,
> Joshua Schaeffer
________________________________________________
Kerberos mailing list Kerberos@mit.edu
https://mailman.mit.edu/mailman/listinfo/kerberos
