[asterisk-bugs] [DAHDI-linux 0013206]: [patch] ztdynamic double buffering on rx, statistics

Asterisk Bug Tracker noreply at bugs.digium.com
Wed Sep 24 03:05:42 CDT 2008


A NOTE has been added to this issue. 
====================================================================== 
http://bugs.digium.com/view.php?id=13206 
====================================================================== 
Reported By:                biohumanoid
Assigned To:                
====================================================================== 
Project:                    DAHDI-linux
Issue ID:                   13206
Category:                   dahdi_dynamic
Reproducibility:            have not tried
Severity:                   major
Priority:                   normal
Status:                     new
====================================================================== 
Date Submitted:             2008-07-31 00:28 CDT
Last Modified:              2008-09-24 03:05 CDT
====================================================================== 
Summary:                    [patch] ztdynamic double buffering on rx, statistics
Description: 
In current implementation:
1. If we have more than one TDMoE device, we'll have a slip/skip. 2 TDMoE
devices can have phase jitter.
2. We have no statistics (which is necessary).
3. ztdynamic can't be a master for zaptel (but why not ?)
====================================================================== 

---------------------------------------------------------------------- 
 (0092820) biohumanoid (reporter) - 2008-09-24 03:05
 http://bugs.digium.com/view.php?id=13206#c92820 
---------------------------------------------------------------------- 
>module_printk(level, fmt, args...) printk(level "%s: " fmt,
THIS_MODULE->name, ## args)

>Actually, this macro is defined twice already. Maybe move it to kernel.h
? 
Yes, you should.

> 1024 as in your comment
artifact from zaptel-1.4...

tasklet statistics takes <= 150 bytes
per-span statistics takes <= 150 bytes

I think 4096/150 - 1 = 26 dynamic spans is much enougth.
i386 can't handle 26000 of interrupts per second...

I've forgot one more bug fixed:
in dahdi_dynamic_unregister, dynamic_destroy(z) can be called with
spinlock..
this finally will call checkmaster with spinlock... 

Issue History 
Date Modified    Username       Field                    Change               
====================================================================== 
2008-09-24 03:05 biohumanoid    Note Added: 0092820                          
======================================================================




More information about the asterisk-bugs mailing list