Summary: | ASTERISK-09742: [patch] ast_debug | ||
Reporter: | Dmitry Andrianov (dimas) | Labels: | |
Date Opened: | 2007-06-24 17:27:12 | Date Closed: | 2007-07-06 19:18:18 |
Priority: | Minor | Regression? | No |
Status: | Closed/Complete | Components: | Core/General |
Versions: | Frequency of Occurrence | ||
Related Issues: | |||
Environment: | Attachments: | ( 0) 9984-missing.patch | |
Description: | Following http://bugs.digium.com/bug_view_advanced_page.php?bug_id=9984 Some of the changes which were in the patch weren't applied. I'm attaching new patch with outstandind modifications. | ||
Comments: | By: Russell Bryant (russell) 2007-06-25 07:43:53 In most of these changes, the check of option_debug is still there. Could you go ahead and remove them? if (blahdebug && option_debug) ast_log(LOG_DEBUG, ...); if (blahdebug) ast_debug(1, ...); By: Dmitry Andrianov (dimas) 2007-06-25 07:57:18 Unfortunately I can not. The check is not (blah && option_debug) but rather (blah || option_debug) so message can be logged even when option_debug is zero but blah is set to true. That is also the reason why I do ast_debug(0, ...) - to make sure message will be logged even is option_debug is zero. I would commit that patch because it at least does not change behaviour. And then think how to improve logging in rtp.c and similar cases. My original idea was to have #define DEBUG_LEVEL option_debug in the logger.h and let .c files to override it like #define DEBUG_LEVEL (rtpdebug ? 100 : option_debug) But this is something everybody should agree first - there are other ways to achieve the same... By: Russell Bryant (russell) 2007-06-25 08:42:08 committed in rev 71557, thanks! |