EZ8 ADAT I/O card (ICE1712 chipset)

EZ8 ADAT I/O card (ICE1712 chipset)

[Doug]

I finally decided to start working on ALSA support for my EZ8 card
(www.event1.com) and man, did I get lucky.  I am brand new to device
driver development altogether, so my main concern here is learning how
everything works in addition to just getting the EZ8 card to work.

Out of the box, the snd-ice1712 driver recognized it as a Hoontech DSP24
card (I also have a Delta 1010LT installed):

cat /proc/asound/cards
0 [M1010LT        ]: ICE1712 - M Audio Delta 1010LT
                      M Audio Delta 1010LT at 0x9000, irq 11
1 [DSP24          ]: ICE1712 - Hoontech SoundTrack Audio DSP24
                      Hoontech SoundTrack Audio DSP24 at 0xa000, irq 11

With this configuration I am able to send PCM out to all 8 channels but
no input.  I refer to my ldd2 book, "Writing an ALSA Driver" document,
and ice1712 src to learn how the cards are identified.  I found that the
  subvendor/subdevice ID @ 0x2c-0x2f are the same between the EZ8 and
the Hoontech DSP24 (0x12141217).  As an experiment I made the following
change to line 32 of hoontech.h:

#define ICE1712_SUBDEVICE_STDSP24		0x12141216

Basically just changing the ID to a bogus number.  As I expected it now
was identified as a generic device:

cat /proc/asound/cards
0 [M1010LT        ]: ICE1712 - M Audio Delta 1010LT
                      M Audio Delta 1010LT at 0x9000, irq 11
1 [ICE1712        ]: ICE1712 - ICEnsemble ICE1712
                      ICEnsemble ICE1712 at 0xa000, irq 11

And bonus!  Everything now works, all 8 channels work in and out and in
full duplex.  So far I have tested 48Khs and 44.1Khz sending 8 channels
up and down at the same time w/ ardour, and envy24control displays all
of the levels accurately.  It appears to be fully functional.  I of
course will be testing it out much more.

Questions:

Is this common for PCI device subid's to overlap?  If so, is there a
preferred method of further identification?  I was looking at the
configuration space for something else that could be used to identify a
device, like maybe the class code area at 0x09-0x0b?

I see some MODULE_PARAM macros in ice1712.c, but I am still unclear on
their exact usage.  Is there already provisions in the driver to force
generic via a module parameter?

I would like to continue further with this to add the EZ8 card to the
list of 'supported' devices for this driver, again mainly for my
interest in learning about device driver development and linux audio.  I
contacted Event about releasing documentation some time ago before I
realized how easy this was going to be, to no avail.  I guess my
question to the ice1712 author is, how is this done?  May I add code
that finds a way to uniquely identifies the EZ8 card and registers it
appropriately?  If I succeed and it meets your aproval, would it be
added to your driver source?

-- 
http://nostar.isa-geek.com/




-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Doug]

Here is what I have done to the ice1712.c file in order to add support 
for my EZ8 card.  It's a temporary hack (emphesis on 'hack') until I add 
code to reliably identify the ez8 from the hoontech (both cards share 
the same subvendorid).

/* EZ8 Hack */
int ez8=0;
MODULE_PARM(ez8, "i");

I added this at the end of all ofthe MODULE_PARM() definitions.

On line 2505, I made this change:

for (tbl = card_tables+(ez8 ? 1 : 0); *tbl; tbl++) { /* EZ8 Hack */

Just a quick and dirty way to skip over the hoontech card_table entry so 
that my EZ8 is recognized as a generic device, while still correctly 
detecting my Delta 1010LT card when ez8=1 is added as argument to the 
snd-ice1712 module.  Module functions normally if ez8 is not added to 
command line (or is set to 0).  This would not work if an ez8 and 
hoontech were used together.

This, of course, is only a temporary fix for my own use, while I work on 
code to correctly differentiate between the ez8 and the hoontech card. 
In addition, there are a few EZ8 specific features that don't seem to be 
implemented with the generic driver.  Since ADAT is fully functional, 
though, I now will spend much more time using the card/driver (with 
ardour/jack) and less finishing support.  I don't think there will be 
any objection here, since I think I am the only linux using EZ8/EZBus 
owner on the planet :)

Does anyone have the ICE1712 chip documentation?  How were the drivers 
for the other ice1712 based cards written?  Were the manufacturers 
contacted and data released under some sort of non-disclosure agreement? 
   I dont get why documentation for a chip is kept so secretive, I mean 
rtl8xxx ethernet tranceiver chip docs, for example, are all over the 
place, that seems like the same thing to me.  I would really like to 
have the documetation so that I can study the ice1712 driver line by 
line and understand what exactly is being done on-board the chip.  I 
just like that kind of stuff, I know there is enough info in the 
existing source for the other cards to figure out how to implement my needs.

-- 
http://nostar.isa-geek.com/



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Takashi Iwai]

At Thu, 05 Feb 2004 03:50:30 -0500,
Doug wrote:
> 
> Here is what I have done to the ice1712.c file in order to add support 
> for my EZ8 card.  It's a temporary hack (emphesis on 'hack') until I add 
> code to reliably identify the ez8 from the hoontech (both cards share 
> the same subvendorid).
> 
> /* EZ8 Hack */
> int ez8=0;
> MODULE_PARM(ez8, "i");
> 
> I added this at the end of all ofthe MODULE_PARM() definitions.
> 
> On line 2505, I made this change:
> 
> for (tbl = card_tables+(ez8 ? 1 : 0); *tbl; tbl++) { /* EZ8 Hack */
> 
> Just a quick and dirty way to skip over the hoontech card_table entry so 
> that my EZ8 is recognized as a generic device, while still correctly 
> detecting my Delta 1010LT card when ez8=1 is added as argument to the 
> snd-ice1712 module.  Module functions normally if ez8 is not added to 
> command line (or is set to 0).  This would not work if an ez8 and 
> hoontech were used together.

i'd suggest to check ez8 option in hoontech.c rather than in the
ice1712.c routine.


> This, of course, is only a temporary fix for my own use, while I work on 
> code to correctly differentiate between the ez8 and the hoontech card. 
> In addition, there are a few EZ8 specific features that don't seem to be 
> implemented with the generic driver.  Since ADAT is fully functional, 
> though, I now will spend much more time using the card/driver (with 
> ardour/jack) and less finishing support.  I don't think there will be 
> any objection here, since I think I am the only linux using EZ8/EZBus 
> owner on the planet :)
> 
> Does anyone have the ICE1712 chip documentation?

ice1712 chip spec is open.  you can find it on ALSA's ftp.
maybe mirrored on somewhere else, too. 

but usually it's not enough for writing the complete driver, since the
choice and connection of codec chips depends on the board.

>   How were the drivers 
> for the other ice1712 based cards written?

in most cases, we have contact with the hardware vendor, and get some
information about the hardware.

>  Were the manufacturers 
> contacted and data released under some sort of non-disclosure agreement? 

depends.


Takashi


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Doug]

Takashi Iwai wrote:
> At Thu, 05 Feb 2004 03:50:30 -0500,
> Doug wrote:
> 
>>Here is what I have done to the ice1712.c file in order to add support 
>>for my EZ8 card.  It's a temporary hack (emphesis on 'hack') until I add 
>>code to reliably identify the ez8 from the hoontech (both cards share 
>>the same subvendorid).
>>
>>/* EZ8 Hack */
>>int ez8=0;
>>MODULE_PARM(ez8, "i");
>>
>>I added this at the end of all ofthe MODULE_PARM() definitions.
>>
>>On line 2505, I made this change:
>>
>>for (tbl = card_tables+(ez8 ? 1 : 0); *tbl; tbl++) { /* EZ8 Hack */
>>
> 
> i'd suggest to check ez8 option in hoontech.c rather than in the
> ice1712.c routine.
> 

Good advice, but after I reealized the chip docs were right there on the 
alsa ftp the whole time, I went ahead and decided to do it right.  I put 
ice1712.c back to its original state, adding a new struct 
snd_ice1712_card_info called snd_ice1712_event_cards, and 
MODULE_DEVICES, etc.  I created an event.c (and .h) and will add support 
following suit with hoontech.c and friends.  Now, armed with the chip 
specs, I will find a way to uniquely identify the ez8 card, and tailor 
the driver to its features.  I'll post the results when finished (or stuck).

Thanks for the info!


-- 
http://nostar.isa-geek.com/



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Doug McLain]

[-- Attachment #1: Type: text/plain, Size: 2894 bytes --]

Here is a patch for EZ8 support with the snd-ice1712 driver.  I kept 
everything contained within hoontech.c, as you suggested.  I never 
bothered to go farther trying to uniquely identify between the 2 cards 
since, this has been working fine for me. Also, I have contacted both 
Hoontech and Event, informing them of the conflict by not creating their 
own uniqe subvendor id. Hopefully one or both will release a driver 
update that will give either one a unique id.  I never made a patch 
before, but it looks fine to me.  Please let me know otherwise.

Just load the driver with the ez8=1 option to enable EZ8 support.

Takashi Iwai wrote:
> At Thu, 05 Feb 2004 03:50:30 -0500,
> Doug wrote:
> 
>>Here is what I have done to the ice1712.c file in order to add support 
>>for my EZ8 card.  It's a temporary hack (emphesis on 'hack') until I add 
>>code to reliably identify the ez8 from the hoontech (both cards share 
>>the same subvendorid).
>>
>>/* EZ8 Hack */
>>int ez8=0;
>>MODULE_PARM(ez8, "i");
>>
>>I added this at the end of all ofthe MODULE_PARM() definitions.
>>
>>On line 2505, I made this change:
>>
>>for (tbl = card_tables+(ez8 ? 1 : 0); *tbl; tbl++) { /* EZ8 Hack */
>>
>>Just a quick and dirty way to skip over the hoontech card_table entry so 
>>that my EZ8 is recognized as a generic device, while still correctly 
>>detecting my Delta 1010LT card when ez8=1 is added as argument to the 
>>snd-ice1712 module.  Module functions normally if ez8 is not added to 
>>command line (or is set to 0).  This would not work if an ez8 and 
>>hoontech were used together.
> 
> 
> i'd suggest to check ez8 option in hoontech.c rather than in the
> ice1712.c routine.
> 
> 
> 
>>This, of course, is only a temporary fix for my own use, while I work on 
>>code to correctly differentiate between the ez8 and the hoontech card. 
>>In addition, there are a few EZ8 specific features that don't seem to be 
>>implemented with the generic driver.  Since ADAT is fully functional, 
>>though, I now will spend much more time using the card/driver (with 
>>ardour/jack) and less finishing support.  I don't think there will be 
>>any objection here, since I think I am the only linux using EZ8/EZBus 
>>owner on the planet :)
>>
>>Does anyone have the ICE1712 chip documentation?
> 
> 
> ice1712 chip spec is open.  you can find it on ALSA's ftp.
> maybe mirrored on somewhere else, too. 
> 
> but usually it's not enough for writing the complete driver, since the
> choice and connection of codec chips depends on the board.
> 
> 
>>  How were the drivers 
>>for the other ice1712 based cards written?
> 
> 
> in most cases, we have contact with the hardware vendor, and get some
> information about the hardware.
> 
> 
>> Were the manufacturers 
>>contacted and data released under some sort of non-disclosure agreement? 
> 
> 
> depends.
> 
> 
> Takashi
> 

--
http://nostar.isa-geek.com/

[-- Attachment #2: alsa-driver-cvs.diff --]
[-- Type: text/plain, Size: 2282 bytes --]

Index: alsa-kernel/pci/ice1712/hoontech.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/pci/ice1712/hoontech.c,v
retrieving revision 1.3
diff -c -r1.3 hoontech.c
*** alsa-kernel/pci/ice1712/hoontech.c  5 Mar 2004 09:08:39 -0000       1.3
--- alsa-kernel/pci/ice1712/hoontech.c  1 Apr 2004 03:31:52 -0000
***************
*** 32,37 ****
--- 32,40 ----
  #include "ice1712.h"
  #include "hoontech.h"

+ /* EZ8 Hack:  pass ez8=1 for EZ8 card */
+ int ez8=0;
+ MODULE_PARM(ez8, "i");

  static void __devinit snd_ice1712_stdsp24_gpio_write(ice1712_t *ice, unsigned char byte)
  {
***************
*** 151,156 ****
--- 154,173 ----
  static int __devinit snd_ice1712_hoontech_init(ice1712_t *ice)
  {
        int box, chn;
+
+       /* EZ8 Hack: Change shortname and subvendor id, Recall functions called in
+        * snd_ice1712_chip_init when it still thinks it is a Hoontech DSP24 card.
+        */
+       if (ez8) {
+               strcpy(ice->card->shortname, "Event Electronics EZ8");
+               ice->eeprom.subvendor = 0;
+               ice->gpio.write_mask = ice->eeprom.gpiomask;
+               ice->gpio.direction = ice->eeprom.gpiodir;
+               snd_ice1712_write(ice, ICE1712_IREG_GPIO_WRITE_MASK, ice->eeprom.gpiomask);
+               snd_ice1712_write(ice, ICE1712_IREG_GPIO_DIRECTION, ice->eeprom.gpiodir);
+               snd_ice1712_write(ice, ICE1712_IREG_GPIO_DATA, ice->eeprom.gpiostate);
+               return 0;
+       }

        ice->num_total_dacs = 8;
        ice->num_total_adcs = 8;
Index: alsa-kernel/pci/ice1712/ice1712.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/pci/ice1712/ice1712.c,v
retrieving revision 1.47
diff -c -r1.47 ice1712.c
*** alsa-kernel/pci/ice1712/ice1712.c   26 Mar 2004 12:25:51 -0000      1.47
--- alsa-kernel/pci/ice1712/ice1712.c   1 Apr 2004 03:31:55 -0000
***************
*** 41,46 ****
--- 41,49 ----
   *  2003.02.20  Taksahi Iwai <tiwai@suse.de>
   *    Split vt1724 part to an independent driver.
   *    The GPIO is accessed through the callback functions now.
+  *
+  * 2004.03.31 Doug McLain <nostar@comcast.net>
+  *    Added support for Event Electronics EZ8 card to hoontech.c.
   */

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Doug McLain]

Whats the word on this patch? Does something need to change?  Is it ok 
as is?

Doug McLain wrote:
> Here is a patch for EZ8 support with the snd-ice1712 driver.  I kept 
> everything contained within hoontech.c, as you suggested.  I never 
> bothered to go farther trying to uniquely identify between the 2 cards 
> since, this has been working fine for me. Also, I have contacted both 
> Hoontech and Event, informing them of the conflict by not creating their 
> own uniqe subvendor id. Hopefully one or both will release a driver 
> update that will give either one a unique id.  I never made a patch 
> before, but it looks fine to me.  Please let me know otherwise.
> 
> Just load the driver with the ez8=1 option to enable EZ8 support.
> 
> Takashi Iwai wrote:
> 
>> At Thu, 05 Feb 2004 03:50:30 -0500,
>> Doug wrote:
>>
>>> Here is what I have done to the ice1712.c file in order to add 
>>> support for my EZ8 card.  It's a temporary hack (emphesis on 'hack') 
>>> until I add code to reliably identify the ez8 from the hoontech (both 
>>> cards share the same subvendorid).
>>>
>>> /* EZ8 Hack */
>>> int ez8=0;
>>> MODULE_PARM(ez8, "i");
>>>
>>> I added this at the end of all ofthe MODULE_PARM() definitions.
>>>
>>> On line 2505, I made this change:
>>>
>>> for (tbl = card_tables+(ez8 ? 1 : 0); *tbl; tbl++) { /* EZ8 Hack */
>>>
>>> Just a quick and dirty way to skip over the hoontech card_table entry 
>>> so that my EZ8 is recognized as a generic device, while still 
>>> correctly detecting my Delta 1010LT card when ez8=1 is added as 
>>> argument to the snd-ice1712 module.  Module functions normally if ez8 
>>> is not added to command line (or is set to 0).  This would not work 
>>> if an ez8 and hoontech were used together.
>>
>>
>>
>> i'd suggest to check ez8 option in hoontech.c rather than in the
>> ice1712.c routine.
>>
>>
>>
>>> This, of course, is only a temporary fix for my own use, while I work 
>>> on code to correctly differentiate between the ez8 and the hoontech 
>>> card. In addition, there are a few EZ8 specific features that don't 
>>> seem to be implemented with the generic driver.  Since ADAT is fully 
>>> functional, though, I now will spend much more time using the 
>>> card/driver (with ardour/jack) and less finishing support.  I don't 
>>> think there will be any objection here, since I think I am the only 
>>> linux using EZ8/EZBus owner on the planet :)
>>>
>>> Does anyone have the ICE1712 chip documentation?
>>
>>
>>
>> ice1712 chip spec is open.  you can find it on ALSA's ftp.
>> maybe mirrored on somewhere else, too.
>> but usually it's not enough for writing the complete driver, since the
>> choice and connection of codec chips depends on the board.
>>
>>
>>>  How were the drivers for the other ice1712 based cards written?
>>
>>
>>
>> in most cases, we have contact with the hardware vendor, and get some
>> information about the hardware.
>>
>>
>>> Were the manufacturers contacted and data released under some sort of 
>>> non-disclosure agreement? 
>>
>>
>>
>> depends.
>>
>>
>> Takashi
>>
> 
> -- 
> http://nostar.isa-geek.com/
> 
> 
> ------------------------------------------------------------------------
> 
> Index: alsa-kernel/pci/ice1712/hoontech.c
> ===================================================================
> RCS file: /cvsroot/alsa/alsa-kernel/pci/ice1712/hoontech.c,v
> retrieving revision 1.3
> diff -c -r1.3 hoontech.c
> *** alsa-kernel/pci/ice1712/hoontech.c  5 Mar 2004 09:08:39 -0000       1.3
> --- alsa-kernel/pci/ice1712/hoontech.c  1 Apr 2004 03:31:52 -0000
> ***************
> *** 32,37 ****
> --- 32,40 ----
>   #include "ice1712.h"
>   #include "hoontech.h"
> 
> + /* EZ8 Hack:  pass ez8=1 for EZ8 card */
> + int ez8=0;
> + MODULE_PARM(ez8, "i");
> 
>   static void __devinit snd_ice1712_stdsp24_gpio_write(ice1712_t *ice, unsigned char byte)
>   {
> ***************
> *** 151,156 ****
> --- 154,173 ----
>   static int __devinit snd_ice1712_hoontech_init(ice1712_t *ice)
>   {
>         int box, chn;
> +
> +       /* EZ8 Hack: Change shortname and subvendor id, Recall functions called in
> +        * snd_ice1712_chip_init when it still thinks it is a Hoontech DSP24 card.
> +        */
> +       if (ez8) {
> +               strcpy(ice->card->shortname, "Event Electronics EZ8");
> +               ice->eeprom.subvendor = 0;
> +               ice->gpio.write_mask = ice->eeprom.gpiomask;
> +               ice->gpio.direction = ice->eeprom.gpiodir;
> +               snd_ice1712_write(ice, ICE1712_IREG_GPIO_WRITE_MASK, ice->eeprom.gpiomask);
> +               snd_ice1712_write(ice, ICE1712_IREG_GPIO_DIRECTION, ice->eeprom.gpiodir);
> +               snd_ice1712_write(ice, ICE1712_IREG_GPIO_DATA, ice->eeprom.gpiostate);
> +               return 0;
> +       }
> 
>         ice->num_total_dacs = 8;
>         ice->num_total_adcs = 8;
> Index: alsa-kernel/pci/ice1712/ice1712.c
> ===================================================================
> RCS file: /cvsroot/alsa/alsa-kernel/pci/ice1712/ice1712.c,v
> retrieving revision 1.47
> diff -c -r1.47 ice1712.c
> *** alsa-kernel/pci/ice1712/ice1712.c   26 Mar 2004 12:25:51 -0000      1.47
> --- alsa-kernel/pci/ice1712/ice1712.c   1 Apr 2004 03:31:55 -0000
> ***************
> *** 41,46 ****
> --- 41,49 ----
>    *  2003.02.20  Taksahi Iwai <tiwai@suse.de>
>    *    Split vt1724 part to an independent driver.
>    *    The GPIO is accessed through the callback functions now.
> +  *
> +  * 2004.03.31 Doug McLain <nostar@comcast.net>
> +  *    Added support for Event Electronics EZ8 card to hoontech.c.
>    */
> 
> 
> 

-- 
http://nostar.isa-geek.com/



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Takashi Iwai]

Doug,

At Fri, 09 Apr 2004 13:59:58 -0400,
Doug McLain wrote:
> 
> Whats the word on this patch? Does something need to change?  Is it ok 
> as is?

sorry i've been too busy for other works for these weeks.
i'll check your patch now.


Takashi


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Takashi Iwai]

At Tue, 13 Apr 2004 12:01:30 +0200,
I wrote:
> 
> Doug,
> 
> At Fri, 09 Apr 2004 13:59:58 -0400,
> Doug McLain wrote:
> > 
> > Whats the word on this patch? Does something need to change?  Is it ok 
> > as is?
> 
> sorry i've been too busy for other works for these weeks.
> i'll check your patch now.

now it's applied to cvs with a slight modification, to make ez8 option
as an array (possibly you can connect multiple ice1712 boards).


thanks for the patch!


Takashi


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: EZ8 ADAT I/O card (ICE1712 chipset)

[Doug McLain]

Groovy...  I have a second EZ8 card sitting unused in a PC at my buddy's 
house.  If he ever takes it out of the PC and brings it here, like he 
has been promising for weeks, I'll experiment with multiple cards.  I'm 
also looking forward to connecting 2 PC's together via the 2 EZ8 cards 
to digitally transfer some cubase 8-track projects to ardour.

Doug

Takashi Iwai wrote:
> At Tue, 13 Apr 2004 12:01:30 +0200,
> I wrote:
> 
>>Doug,
>>
>>At Fri, 09 Apr 2004 13:59:58 -0400,
>>Doug McLain wrote:
>>
>>>Whats the word on this patch? Does something need to change?  Is it ok 
>>>as is?
>>
>>sorry i've been too busy for other works for these weeks.
>>i'll check your patch now.
> 
> 
> now it's applied to cvs with a slight modification, to make ez8 option
> as an array (possibly you can connect multiple ice1712 boards).
> 
> 
> thanks for the patch!
> 
> 
> Takashi
> 

-- 
http://nostar.isa-geek.com/



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

EZ8 ADAT I/O card (ICE1712 chipset)

[Doug]

I finally decided to start working on ALSA support for my EZ8 card 
(www.event1.com) and man, did I get lucky.  I am brand new to device 
driver development altogether, so my main concern here is learning how 
everything works in addition to just getting the EZ8 card to work.

Out of the box, the snd-ice1712 driver recognized it as a Hoontech DSP24 
card (I also have a Delta 1010LT installed):

cat /proc/asound/cards
0 [M1010LT        ]: ICE1712 - M Audio Delta 1010LT
                      M Audio Delta 1010LT at 0x9000, irq 11
1 [DSP24          ]: ICE1712 - Hoontech SoundTrack Audio DSP24
                      Hoontech SoundTrack Audio DSP24 at 0xa000, irq 11

With this configuration I am able to send PCM out to all 8 channels but 
no input.  I refer to my ldd2 book, "Writing an ALSA Driver" document, 
and ice1712 src to learn how the cards are identified.  I found that the 
  subvendor/subdevice ID @ 0x2c-0x2f are the same between the EZ8 and 
the Hoontech DSP24 (0x12141217).  As an experiment I made the following 
change to line 32 of hoontech.h:

#define ICE1712_SUBDEVICE_STDSP24		0x12141216

Basically just changing the ID to a bogus number.  As I expected it now 
was identified as a generic device:

cat /proc/asound/cards
0 [M1010LT        ]: ICE1712 - M Audio Delta 1010LT
                      M Audio Delta 1010LT at 0x9000, irq 11
1 [ICE1712        ]: ICE1712 - ICEnsemble ICE1712
                      ICEnsemble ICE1712 at 0xa000, irq 11

And bonus!  Everything now works, all 8 channels work in and out and in 
full duplex.  So far I have tested 48Khs and 44.1Khz sending 8 channels 
up and down at the same time w/ ardour, and envy24control displays all 
of the levels accurately.  It appears to be fully functional.  I of 
course will be testing it out much more.

Questions:

Is this common for PCI device subid's to overlap?  If so, is there a 
preferred method of further identification?  I was looking at the 
configuration space for something else that could be used to identify a 
device, like maybe the class code area at 0x09-0x0b?

I see some MODULE_PARAM macros in ice1712.c, but I am still unclear on 
their exact usage.  Is there already provisions in the driver to force 
generic via a module parameter?

I would like to continue further with this to add the EZ8 card to the 
list of 'supported' devices for this driver, again mainly for my 
interest in learning about device driver development and linux audio.  I 
contacted Event about releasing documentation some time ago before I 
realized how easy this was going to be, to no avail.  I guess my 
question to the ice1712 author is, how is this done?  May I add code 
that finds a way to uniquely identifies the EZ8 card and registers it 
appropriately?  If I succeed and it meets your aproval, would it be 
added to your driver source?

-- 
http://nostar.isa-geek.com/



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn