Summary:ASTERISK-00454: [patch] Unload modules at shutdown/restart
Reporter:Paul Cadach (pcadach)Labels:
Date Opened:2003-10-29 14:17:06.000-0600Date Closed:2011-06-07 14:10:12
Versions:Frequency of
Environment:Attachments:( 0) asterisk-0.5.0-unload.diff
( 1) asterisk-0.5.0-unload2.diff
( 2) cvs-unload-all.diff
Description:Included patch contains 3 "independed" parts:
1) module unloader (registers itself for ast_atexit() to not patch asterisk.c);
2) patches to cleanly unregister ast_cli's;
3) patch to cancel monitor thread in chan_h323.


Works for me (at least). Built on top of asterisk-0.5.0 installation.
Comments:By: Paul Cadach (pcadach) 2003-11-02 13:19:54.000-0600

More things to come (0.5.0-unload2.diff) - found when debugged under RH-7.0 (with glibc-2.2-9, under RH-7.3 with glibc-2.3.2 all works without patching).

By: Paul Cadach (pcadach) 2003-11-17 11:54:52.000-0600

Does those patches will be applied?

By: izo (izo) 2003-11-17 17:35:16.000-0600

Even if they want be applied to main tree thank you for your contribution.

By: Brian West (bkw918) 2003-11-19 19:15:39.000-0600

does this apply to latest cvs?

By: Paul Cadach (pcadach) 2003-12-02 22:43:53.000-0600

Cumulative patch for current cvs attached (unload-all.diff). Includes io and sched cleanups...

By: jerjer (jerjer) 2003-12-23 16:44:52.000-0600

does this overlap with Bug 469?

By: Paul Cadach (pcadach) 2003-12-23 20:30:25.000-0600

No. This ticket is for other modules than H.323 while ASTERISK-465 is for H.323 only.

By: jrollyson (jrollyson) 2004-01-28 00:01:20.000-0600

Unloading modules is a hack. Modules should handle a reload cleanly, if they don't, bugs should be filed on them, with details on how they crash, and backtraces if possible.