[asterisk-dev] [Code Review] a driver for single-port FXO cards based on Si3052 chip + Si3011/17/18 DAA (Motorola 52-6116-2A, PM560MS etc.)

Tzafrir Cohen tzafrir.cohen at xorcom.com
Thu Oct 21 07:38:56 CDT 2010


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviewboard.asterisk.org/r/977/#review2850
-----------------------------------------------------------


Again, the minor nit about indentation: the checkpatch script gives them because you indent your code with 4 spaces, whereas kernel code is indented with tabs that are intended to be 8 chars. Ratinale: Linus[1] :-)

[1] http://kernel.org/doc/Documentation/CodingStyle


/linux/trunk/drivers/dahdi/fxo_si3052.h
<https://reviewboard.asterisk.org/r/977/#comment6030>

    Have you looked at the debug macros in include/dahdi/kernel.h ? The good thing about your macros is that they don't have any run-time overhead if not enabled. The bad thing is that the driver needs to be rebuilt to use them.



/linux/trunk/drivers/dahdi/fxo_si3052.c
<https://reviewboard.asterisk.org/r/977/#comment6029>

    If I have more than one device, and I see that message in my logs, how can I tell which of those devices has generated the message?


- Tzafrir


On 2010-10-19 12:20:36, mhej wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviewboard.asterisk.org/r/977/
> -----------------------------------------------------------
> 
> (Updated 2010-10-19 12:20:36)
> 
> 
> Review request for Asterisk Developers.
> 
> 
> Summary
> -------
> 
> Attached here is a driver for single-port FXO cards based on Si3052 chip + Si3011/17/18 DAA. They are marketed as Motorola 52-6116-2A, PM560MS and many others.
> Driver is written from scratch to take advantage of international features some of the DAAs have. There is also patch for wcfxo module to support these cards circulating on Net, but it runs the card in backwards-compatible mode and doesn't allow most of the POTS settings to be tweaked. 
> 
> Usage:
> * DAHDI-tools need a simple patch from bug https://issues.asterisk.org/view.php?id=18138 (dahdi-gencfg.patch) for dahdi_genconf to recognize new driver.
> * You have to set the region name to one of those listed in fxo_modes.h file (this file is also used for wctdm driver, so region names are the same). For example: "modprobe fxo_si3052 region_param=ARGENTINA" (w/o quotes) would load the module with proper settings for Argentina. Also add "use_ulaw=1" if you plan to use ULAW instead of ALAW (default setting is ALAW).
> * Generate config file for the card with dahdi_genconf, configure with dahdi_cfg, write config files for Asterisk, all just like with the wcfxo driver.
> 
> If CallerID detection doesn't work the problem might be with too long minimum ringer time. Default is 512ms, but sometimes first ring signal is shorter and is not detected as ring, so CallerID detection is not started by Asterisk. In this case check ringer length and add appropriate "min_ringer_time" parameter (for example "min_ringer_time=150" for 150ms). Valid values are 100, 150, 200, 256, 384, 512 and 1024 ms.
>  
> Comments and suggestions are welcome.
> 
> 
> This addresses bug 18138.
>     http://issues.asterisk.org/view.php?id=18138
> 
> 
> Diffs
> -----
> 
>   /linux/trunk/drivers/dahdi/Kbuild 9431 
>   /linux/trunk/drivers/dahdi/fxo_modes.h 9431 
>   /linux/trunk/drivers/dahdi/fxo_si3052.h PRE-CREATION 
>   /linux/trunk/drivers/dahdi/fxo_si3052.c PRE-CREATION 
> 
> Diff: https://reviewboard.asterisk.org/r/977/diff
> 
> 
> Testing
> -------
> 
> Tested features:
> * Sending and receiving voice,
> * Ringer detection and validation,
> * DTMF detection,
> * Caller ID reception (only ETSI FSK mode tested, but different ones should work too in principle),
> * ULAW/ALAW mode,
> * Busy pattern hangup detecion,
> * Line disconnection detection (only when offhook due to hardware limitation).
> * No errors were indicated by checkpatch.pl.
> 
> 
> Thanks,
> 
> mhej
> 
>




More information about the asterisk-dev mailing list