| June 23, 2003 | ||
|---|---|---|
 |
View be6f6ec9bf
ser made more noisy: default log level lifted to L_NOTICEJiri Kuthan authored on 23/06/2003 22:13:10 |
|
|
View 4783be5692
avoid abort on unsuccessful module startupJiri Kuthan authored on 23/06/2003 15:43:55 |
|
| June 11, 2003 | ||
|---|---|---|
 |
View 428e3b83cf
- signals handlers are installed after daemonize to avoid catching our own SIGCHLD generated when becoming session leader; thanks to Daniel for diagnosticating this on RH 9.Andrei Pelinescu-Onciul authored on 11/06/2003 13:49:44 |
|
| May 28, 2003 | ||
|---|---|---|
|
View 6b32177f2c
- changed memlog to L_INFO (so that mem. exit dumps will be displayed also on default log level (3)). Other malloc debuging msgs. are still at L_DBG.Andrei Pelinescu-Onciul authored on 28/05/2003 16:11:45 |
|
| May 9, 2003 | ||
|---|---|---|
|
View 34592dd540
- fixed logging to syslog on solaris (on solaris openlog does not reopen the logs fd even if it's closed [daemonize], closelog must be called before, of course this is not documented anywhere).Andrei Pelinescu-Onciul authored on 09/05/2003 16:21:14 |
|
| April 18, 2003 | ||
|---|---|---|
|
View 200c02768d
- fixed process_count() (forgot to add some() when I added tcp_disable)Andrei Pelinescu-Onciul authored on 18/04/2003 13:58:15 |
|
| April 16, 2003 | ||
|---|---|---|
|
View 0762e9d530
- removed all the experimental modules from make modules (there are still some non-experimantal ones removed because of library dependecies)Andrei Pelinescu-Onciul authored on 16/04/2003 18:22:47 |
|
| April 15, 2003 | ||
|---|---|---|
|
View 5dcfb23d56
- added disable tcp support (-T or disable_tcp=yes in the cfg) - the number of tcp "worker" processes is now setable (-N or tcp_children=NN) - replaced get_out_socket w/ get_send_socket in tm/uac.c & fixed some proto stuff (proto was not set properly when called w/ PROTO_NONE) - updated man pages & INSTALLAndrei Pelinescu-Onciul authored on 15/04/2003 20:39:37 |
|
| April 10, 2003 | ||
|---|---|---|
|
View 45a03e3cf7
- sendipv[46] was not set in no fork mode (found by Jan)Andrei Pelinescu-Onciul authored on 10/04/2003 22:06:30 |
|
| April 8, 2003 | ||
|---|---|---|
|
View 90907c3f1e
- split init_mallocs into init_pkg_mallocs & init_shm_mallocs - init_shm_mallocs called after cmd. line parsing (so -m is not ignored any more)Andrei Pelinescu-Onciul authored on 08/04/2003 16:02:43 |
|
 |
View 50d7560030
- forgot to initialize the timer process when forking is enabledJan Janak authored on 08/04/2003 12:31:40 |
|
| April 6, 2003 | ||
|---|---|---|
|
View cb87691afc
- child init called in all ser processes (main, timer, fifo, tcp_main, udp_children, tcp_children, tcp_main) - defined macros PROC_MAIN, PROC_TIMER, PROC_FIFO, PROC_TCP_MAIN - they substitute rank for special processes - listening children have rank 1 and higherJan Janak authored on 06/04/2003 20:28:12 |
|
|
View 4c4e112f78
- timer will use first ipv4/ipv6 socket for sending and not the first interface (which can be a loopback) - rport related src_port fixes (all ports are now kept in host byte order)Andrei Pelinescu-Onciul authored on 06/04/2003 00:21:13 |
|
| April 2, 2003 | ||
|---|---|---|
|
View 1971054445
- fixed ":" in port_no_str bug. Now port_no_str no longer contains the colon (updated sms & tm)Andrei Pelinescu-Onciul authored on 02/04/2003 15:51:03 |
|
| March 31, 2003 | ||
|---|---|---|
|
View 7ec958f3ac
- removed snmp relicts from core (main.c) - init_child is now called also for tcp receivers - tcp: bind_address is properly set to the correct address family or 0 if not init. (now bind_address will point to the first tcp ${AF} listen address -- to fix someday)Andrei Pelinescu-Onciul authored on 31/03/2003 12:28:09 |
|