[Home]

Summary:ASTERISK-11651: Use ast_verb instead of ast_debug for channel driver debugging
Reporter:dea (dea)Labels:
Date Opened:2008-03-16 01:44:31Date Closed:2008-04-09 13:27:37
Priority:MinorRegression?No
Status:Closed/CompleteComponents:Channels/chan_skinny
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) chan_skinny-debug-log.txt
Description:When turning on channel driver specific debugging (skinny set debug on),
it is cleaner to use ast_verb instead of ast_debug.

With core verbose set to 1, there is not a lot of output to the log
or console, but when core debug is set to 1 the logs can get quite
crowded.

MVanbaak-
  I do not believe I can assign reports/tickets, so I leave it to
one of our administrators or bug marshalls to assign it to you.
Comments:By: Tilghman Lesher (tilghman) 2008-03-16 10:12:38

What's wrong with using the file-specific debugging, as in:

CLI> core set debug 5 chan_skinny.so

This turns on debugging ONLY for that one file, which is probably what you're aiming to see.
By: dea (dea) 2008-03-16 10:57:09

That almost works.  The majority of the debug logging in chan_skinny,
as well as a number of other files, happen to be protected by channel/app
specific debug commands.  So it would take two cli commands to produce
any output.

This patch semi-reverts one of the changes made when all logging was
converted to the ast_verb and ast_debug macros.  Prior to that conversion
the majority of chan_skinny debug output was protected by 'skinny set debug'
and used ast_verbose.  After the conversion the same protected logging
statements were now ast_debug.
By: Tilghman Lesher (tilghman) 2008-03-16 11:11:17

Instead of doing that, I'd recommend that you remove the skinny_debug flag, so that there's a COMMON interface for debugging for all files, rather than doing something different for skinny.
By: dea (dea) 2008-03-16 12:16:39

Common?

Channels that use a local debug variable & ast_verb (15%)
IAX2                MGCP
Unistim

Channels that use a local debug variable & ast_debug: (15%)
H323                SIP
Skinny (used to be verbose)

Channels that use a local debug and raw CLI output (15%)
MISDN               OSS (uses ast_log)
USBRadio (ast_log)

Channels that use ast_debug alone (55%)
gtalk               Agent
ALSA                Console
Features            Jingle
NBS                 Phone
VPB                 ZAP
Local

To make it even more interesting, I would not want to just
strip out the skinny_debug flag, I would want to try to assign
each message to a valid level.  So I gave a quick look at the
other channels use of ast_debug.

67% of all channel debuging is at level 1
13% is at level 3
13% is at level 4
7%  is at level 2
.1% is at level 5


I know that it has been discussed in the past to try
defining what type of reporting is desired at each level,
but I have not located any documentation to suggest a
concensus was reached.

In any case I guess the only way forward would be to stip
skinny_debug and leave everything at debug level 1 for now
so that we can be moving towards the common interface.
Even so, chan_skinny is far from unique in this regard.
By: Digium Subversion (svnbot) 2008-04-09 12:36:23

Repository: asterisk
Revision: 113834

U   trunk/channels/chan_skinny.c

------------------------------------------------------------------------
r113834 | qwell | 2008-04-09 12:36:19 -0500 (Wed, 09 Apr 2008) | 7 lines

Move all messages wrapped in skinnydebug from debug to verbose.

(closes issue ASTERISK-11651)
Reported by: DEA
Patches:
     chan_skinny-debug-log.txt uploaded by DEA (license 3)

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=113834
By: Digium Subversion (svnbot) 2008-04-09 12:36:50

Repository: asterisk
Revision: 113835

_U  branches/1.6.0/

------------------------------------------------------------------------
r113835 | qwell | 2008-04-09 12:36:50 -0500 (Wed, 09 Apr 2008) | 14 lines

Blocked revisions 113834 via svnmerge

........
r113834 | qwell | 2008-04-09 12:41:09 -0500 (Wed, 09 Apr 2008) | 7 lines

Move all messages wrapped in skinnydebug from debug to verbose.

(closes issue ASTERISK-11651)
Reported by: DEA
Patches:
     chan_skinny-debug-log.txt uploaded by DEA (license 3)

........

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=113835
By: Digium Subversion (svnbot) 2008-04-09 13:27:37

Repository: asterisk
Revision: 113872

_U  team/group/codec_bits/
U   team/group/codec_bits/channels/chan_h323.c
U   team/group/codec_bits/channels/chan_skinny.c
U   team/group/codec_bits/contrib/scripts/astcli
U   team/group/codec_bits/main/pbx.c

------------------------------------------------------------------------
r113872 | tilghman | 2008-04-09 13:27:35 -0500 (Wed, 09 Apr 2008) | 44 lines

Merged revisions 113834,113836,113838,113840 via svnmerge from
https://origsvn.digium.com/svn/asterisk/trunk

........
r113834 | qwell | 2008-04-09 12:41:09 -0500 (Wed, 09 Apr 2008) | 7 lines

Move all messages wrapped in skinnydebug from debug to verbose.

(closes issue ASTERISK-11651)
Reported by: DEA
Patches:
     chan_skinny-debug-log.txt uploaded by DEA (license 3)

........
r113836 | mmichelson | 2008-04-09 12:48:33 -0500 (Wed, 09 Apr 2008) | 14 lines

There was a subtle logical difference between 1.4 and trunk with regards to how timeouts
were handled. In 1.4, if the absolute timeout were reached on a call, no matter what
the return value of ast_spawn_extension was, the pbx would attempt to go to the 'T'
extension or hangup otherwise. The rearrangement of this function in trunk made this check
only happen in the case that ast_spawn_extension returned 0. If ast_spawn_extension returned
1, then the fact that the timeout expired resulted in a no-op, and would cause an infinite
loop to occur in __ast_pbx_run. This change fixes this problem. Now timeouts will
behave as they did in 1.4

(closes issue ASTERISK-11038)
Reported by: pj
Tested by: putnopvut


........
r113838 | qwell | 2008-04-09 12:56:07 -0500 (Wed, 09 Apr 2008) | 2 lines

Fix a small file handle "leak" pointed out by jjshoe on #asterisk.

........
r113840 | file | 2008-04-09 13:05:40 -0500 (Wed, 09 Apr 2008) | 4 lines

Enable enough RTP bridging to allow P2P to work.
(closes issue ASTERISK-11355)
Reported by: pj

........

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=113872