[Dovecot] How to create managesieve core dumps
I've found this in my dovecot.log: ,--[ /path/to/dovecot.log ]-- | Jul 11 01:44:29 dovecot: Info: Dovecot v1.2.1 starting up | ... | Jul 11 01:50:21 auth(default): Info: client in: AUTH 1 PLAIN service=sieve lip=192.168.111.222 rip=192.168.111.122 lport=12000 rport=35084 resp=<hidden> | Jul 11 01:50:21 auth(default): Info: sql(j.doe@example.com,192.168.111.122): query: SELECT "user", password FROM dovecot_password WHERE "user" = 'j.doe@example.com' AND sieve | Jul 11 01:50:21 auth(default): Info: client out: OK 1 user=j.doe@example.com | Jul 11 01:50:21 auth(default): Info: master in: REQUEST 1 31573 1 | Jul 11 01:50:21 auth(default): Info: sql(j.doe@example.com,192.168.111.122): SELECT home, uid, gid, 'dbox:~/dbox' AS mail FROM dovecot_user WHERE userid = 'j.doe@example.com' | Jul 11 01:50:21 auth(default): Info: master out: USER 1 j.doe@example.com home=/srv/mail/9/70037/70035 uid=70035 gid=70037 mail=dbox:~/dbox ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | Jul 11 01:50:21 managesieve-login: Info: Login: user=j.doe@example.com, method=PLAIN, rip=192.168.111.122, lip=192.168.111.222 | Jul 11 01:50:21 auth(default): Info: new auth connection: pid=370 | Jul 11 01:51:25 MANAGESIEVE(j.doe@example.com): Panic: pool_data_stack_realloc(): stack frame changed | Jul 11 01:51:25 MANAGESIEVE(j.doe@example.com): Error: Raw backtrace: managesieve [0x4a45da] -> managesieve [0x4a5360] -> managesieve(i_fatal+0) [0x4a48ce] -> managesieve [0x4b1e6a] -> managesieve [0x4a2421] -> managesieve [0x4a28e9] -> managesieve(buffer_get_space_unsafe+0x2d) [0x4a2d0b] -> managesieve(str_vprintfa+0x6c) [0x4bb149] -> managesieve [0x42a9f7] -> managesieve(sieve_verror+0x100) [0x429c26] -> managesieve [0x44f0a5] -> managesieve [0x44eedf] -> managesieve(sieve_lexer_skip_token+0x15) [0x44f0c8] -> managesieve [0x4322be] -> managesieve [0x4317b2] -> managesieve [0x431ec0] -> managesieve(sieve_parser_run+0x95) [0x432152] -> managesieve(sieve_parse+0x36) [0x42b44b] -> managesieve(sieve_compile_script+0x1d) [0x42b538] -> managesieve [0x4210f5] -> managesieve [0x421678] -> managesieve [0x421474] -> managesieve [0x421787] -> managesieve(cmd_putscript+0x48) [0x4217d1] -> managesieve [0x4231b8] -> managesieve [0x423365] -> managesieve(client_input+0x100) [0x42346f] -> managesieve(io_loop_handler_run+0x17d) [0x4aecfb] | Jul 11 01:51:25 dovecot: Error: child 366 (managesieve) killed with signal 6 (core not dumped - is home dir set?) `-- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Dovecot v1.2.1/f353c5b71097 was started with: host ~ # ulimit -c unlimited hots ~ # /usr/local/dovecot12/sbin/dovecot ILoading modules from directory: ... IEffective uid=65534, gid=65534, home=/tmp ... ^^^^^^^^^ IEffective uid=65534, gid=65534, home=/tmp ^^^^^^^^^
Why is Dovecot not dumping a core and asking: »is home dir set?«?
Regards, Pascal
The trapper recommends today: cafebabe.0919315@localdomain.org
participants (1)
-
Pascal Volk