[Home]

Summary:ASTERISK-00660: using soft hangup zap emits "Use STOP NOW to shutdown Asterisk"
Reporter:Andrew Kohlsmith (akohlsmith)Labels:
Date Opened:2003-12-12 08:10:21.000-0600Date Closed:2004-09-25 02:55:31
Priority:MinorRegression?No
Status:Closed/CompleteComponents:Core/General
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:
Description:*CLI> soft hangup Zap/17-1
Requested Hangup on channel 'Zap/17-1'
   -- Hungup 'Zap/13-1'
 == Spawn extension (in_2914574, s, 2) exited non-zero on 'Zap/17-1'
   -- Executing Hangup("Zap/17-1", "") in new stack
 == Spawn extension (in_2914574, h, 1) exited non-zero on 'Zap/17-1'
   -- Hungup 'Zap/17-1'

Use STOP NOW to shutdown Asterisk
*CLI>

I'm pretty sure that "Use STOP NOW to shutdown Asterisk" is not supposed to be there.  :-)

****** ADDITIONAL INFORMATION ******

AKA satan bug!
Comments:By: Brian West (bkw918) 2003-12-12 13:52:06.000-0600

Doesn't do it here.. what cvs are you running?

By: Brian West (bkw918) 2003-12-12 14:08:07.000-0600

only shows up when you run asterisk -vvvvgc doesn't show up in asterisk -r... hrm strange.

By: Brian West (bkw918) 2003-12-12 14:24:42.000-0600

I'm thinking that else shouldn't be there.  Because it could in theory pop up that error message at other times when running like that.

By: Tilghman Lesher (tilghman) 2003-12-12 21:40:00.000-0600

I disagree.  The code is correct, but there's something very odd going on here.  el_gets() should not return NULL in this code, unless some control character is getting lodged somehow into the readline buffer.

I'd say this is a very insidious bug, having to do with a corrupted input buffer and needs to be tracked down.  It's lucky this didn't cause a crash!

By: Brian West (bkw918) 2003-12-12 23:35:46.000-0600

that message will always appear if option_remote = null right?  or am I reading it wrong?

By: Tilghman Lesher (tilghman) 2003-12-13 09:05:17.000-0600

That's simply asking "Am I in console mode or remote mode for this error?"  Further up is where the actual problem is, the return value of el_gets(), which is supposed to get a line of input.  What this is saying is that there was a line of garbage waiting for el_gets(), probably containing a Ctrl-D (EOF), which caused el_gets() to signal an error by returning NULL.

The purpose of this portion of code is to remind people that they are on the console of Asterisk and quitting has serious implications (i.e. shutting down the phone system).

The fact that 'soft hangup zap/nn' put garbage into the readline buffer is the real problem, which the patch uploaded does nothing to fix (it simply removes a symptom of the problem).

By: maik (maik) 2003-12-29 21:33:10.000-0600

I've just found a bug that seems to be related to this one. I had * running in console-mode and did a 'restart when convenient' from a remote console. After that * repeated the 'Use STOP ...' message over and over again on the console.

By: zoa (zoa) 2004-02-25 06:33:35.000-0600

that last bug is fixed.

But is the 666 bug fixed ?

By: Andrew Kohlsmith (akohlsmith) 2004-02-25 10:55:41.000-0600

Yes bug 666 appears to be fixed now.  I just updated my * sources and with -vvvvgc it does not emit a "use STOP NOW" message when I soft hangup a Zap channel.

Who fixed the bug?

By: zoa (zoa) 2004-02-25 11:41:28.000-0600

i think it got fixed together with the bugfix for the repeating STOP NOW flood on asterisk -r stop now.

By: zoa (zoa) 2004-02-25 11:45:02.000-0600

fixed by james together with :

http://bugs.digium.com/bug_view_page.php?bug_id=0000864

bugfix also confirmed by tzanger.