RME DIGI96 ADAT input selection bug fix

RME DIGI96 ADAT input selection bug fix

[Anders Torger]

Stefan Mueller found a bug in my Digi96 driver. Sometimes I write != 
when I mean == or the other way around:

--- rme96.old.c 2003-07-01 20:13:40.000000000 +0200
+++ rme96.new.c 2003-07-01 20:13:26.000000000 +0200
@@ -1318,7 +1318,7 @@
        snd_pcm_set_sync(substream);

        runtime->hw = snd_rme96_capture_adat_info;
-        if (snd_rme96_getinputtype(rme96) != RME96_INPUT_ANALOG) {
+        if (snd_rme96_getinputtype(rme96) == RME96_INPUT_ANALOG) {
                 /* makes no sense to use analog input. Note that analog
                    expension cards AEB4/8-I are RME96_INPUT_INTERNAL */
                 return -EIO;



-------------------------------------------------------
This SF.Net email sponsored by: Free pre-built ASP.NET sites including
Data Reports, E-commerce, Portals, and Forums are available now.
Download today and enter to win an XBOX or Visual Studio .NET.
http://aspnet.click-url.com/go/psa00100006ave/direct;at.asp_061203_01/01

Re: RME DIGI96 ADAT input selection bug fix

[Takashi Iwai]

Hi,

At Tue, 1 Jul 2003 20:14:59 +0200,
Anders Torger wrote:
> 
> 
> Stefan Mueller found a bug in my Digi96 driver. Sometimes I write != 
> when I mean == or the other way around:

thanks, applied to cvs.

BTW, shall i apply the patch by Martin Langer for unifying the clock
source control?  it's been pending for long time for evaluation.

first of all, the patch is needed for rmedigicontrol utility, which
i'd like to put to alsa-tolls tree now.
it's not critical at all, and nice to have the same controls for all
rme cards.


Takashi


-------------------------------------------------------
This SF.Net email sponsored by: Free pre-built ASP.NET sites including
Data Reports, E-commerce, Portals, and Forums are available now.
Download today and enter to win an XBOX or Visual Studio .NET.
http://aspnet.click-url.com/go/psa00100006ave/direct;at.asp_061203_01/01

Re: RME DIGI96 ADAT input selection bug fix

[Anders Torger]

On Wednesday 02 July 2003 12.17, Takashi Iwai wrote:
> Hi,
>
> At Tue, 1 Jul 2003 20:14:59 +0200,
>
> Anders Torger wrote:
> > Stefan Mueller found a bug in my Digi96 driver. Sometimes I write
> > != when I mean == or the other way around:
>
> thanks, applied to cvs.
>
> BTW, shall i apply the patch by Martin Langer for unifying the clock
> source control?  it's been pending for long time for evaluation.

Oh, sorry, I have forgot about that one, could I get to look at the 
patch again? I suppose it is ok, but it could potentially make the 
driver refuse sampling from the input (returning -EIO) when there is no 
signal there. The driver was like that previously, but I don't think 
that is a good behaviour, so I changed it.

/Anders



-------------------------------------------------------
This SF.Net email sponsored by: Free pre-built ASP.NET sites including
Data Reports, E-commerce, Portals, and Forums are available now.
Download today and enter to win an XBOX or Visual Studio .NET.
http://aspnet.click-url.com/go/psa00100006ave/direct;at.asp_061203_01/01

Re: RME DIGI96 ADAT input selection bug fix

[Takashi Iwai]

[-- Attachment #1: Type: text/plain, Size: 1022 bytes --]

At Wed, 2 Jul 2003 15:03:28 +0200,
Anders Torger wrote:
> 
> 
> On Wednesday 02 July 2003 12.17, Takashi Iwai wrote:
> > Hi,
> >
> > At Tue, 1 Jul 2003 20:14:59 +0200,
> >
> > Anders Torger wrote:
> > > Stefan Mueller found a bug in my Digi96 driver. Sometimes I write
> > > != when I mean == or the other way around:
> >
> > thanks, applied to cvs.
> >
> > BTW, shall i apply the patch by Martin Langer for unifying the clock
> > source control?  it's been pending for long time for evaluation.
> 
> Oh, sorry, I have forgot about that one, could I get to look at the 
> patch again? I suppose it is ok, but it could potentially make the 
> driver refuse sampling from the input (returning -EIO) when there is no 
> signal there.

well, i think you're referring to a different patch (yeah, i remember
there was something like that, too).

the patch below is the one i wanted to include.
basically this changes only the name of clock source control element
to be compliant to the definition in ControlNames.txt.


Takashi

[-- Attachment #2: rme96.diff --]
[-- Type: text/plain, Size: 2162 bytes --]

--- rme96.c	Wed May  7 22:24:15 2003
+++ rme96.c.PATCHED	Sat May  3 19:34:45 2003
@@ -806,10 +806,12 @@
 {
 	switch (mode) {
 	case RME96_CLOCKMODE_SLAVE:
+	        /* AutoSync */ 
 		rme96->wcreg &= ~RME96_WCR_MASTER;
 		rme96->areg &= ~RME96_AR_WSEL;
 		break;
 	case RME96_CLOCKMODE_MASTER:
+	        /* Internal */
 		rme96->wcreg |= RME96_WCR_MASTER;
 		rme96->areg &= ~RME96_AR_WSEL;
 		break;
@@ -1862,15 +1864,15 @@
 		snd_iprintf(buffer, "  sample format: 16 bit\n");
 	}
 	if (rme96->areg & RME96_AR_WSEL) {
-		snd_iprintf(buffer, "  clock mode: word clock\n");
+		snd_iprintf(buffer, "  sample clock source: word clock\n");
 	} else if (rme96->wcreg & RME96_WCR_MASTER) {
-		snd_iprintf(buffer, "  clock mode: master\n");
+		snd_iprintf(buffer, "  sample clock source: internal\n");
 	} else if (snd_rme96_getinputtype(rme96) == RME96_INPUT_ANALOG) {
-		snd_iprintf(buffer, "  clock mode: slave (master anyway due to analog input setting)\n");
+		snd_iprintf(buffer, "  sample clock source: autosync (internal anyway due to analog input setting)\n");
 	} else if (snd_rme96_capture_getrate(rme96, &n) < 0) {
-		snd_iprintf(buffer, "  clock mode: slave (master anyway due to no valid signal)\n");
+		snd_iprintf(buffer, "  sample clock source: autosync (internal anyway due to no valid signal)\n");
 	} else {
-		snd_iprintf(buffer, "  clock mode: slave\n");
+		snd_iprintf(buffer, "  sample clock source: autosync\n");
 	}
 	if (rme96->wcreg & RME96_WCR_PRO) {
 		snd_iprintf(buffer, "  format: AES/EBU (professional)\n");
@@ -2095,7 +2097,7 @@
 static int
 snd_rme96_info_clockmode_control(snd_kcontrol_t *kcontrol, snd_ctl_elem_info_t * uinfo)
 {
-	static char *texts[3] = { "Slave", "Master", "Wordclock" };
+	static char *texts[3] = { "AutoSync", "Internal", "Word" };
 	
 	uinfo->type = SNDRV_CTL_ELEM_TYPE_ENUMERATED;
 	uinfo->count = 1;
@@ -2418,7 +2420,7 @@
 },
 {
         .iface =        SNDRV_CTL_ELEM_IFACE_PCM,
-	.name =         "Clock Mode",
+	.name =         "Sample Clock Source",
 	.info =         snd_rme96_info_clockmode_control, 
 	.get =          snd_rme96_get_clockmode_control,
 	.put =          snd_rme96_put_clockmode_control

Re: RME DIGI96 ADAT input selection bug fix

[Anders Torger]

On Wed, 2 Jul 2003, Takashi Iwai wrote:

> At Wed, 2 Jul 2003 15:03:28 +0200,
> Anders Torger wrote:
> >
> >
> > On Wednesday 02 July 2003 12.17, Takashi Iwai wrote:
> > > Hi,
> > >
> > > At Tue, 1 Jul 2003 20:14:59 +0200,
> > >
> > > Anders Torger wrote:
> > > > Stefan Mueller found a bug in my Digi96 driver. Sometimes I write
> > > > != when I mean == or the other way around:
> > >
> > > thanks, applied to cvs.
> > >
> > > BTW, shall i apply the patch by Martin Langer for unifying the clock
> > > source control?  it's been pending for long time for evaluation.
> >
> > Oh, sorry, I have forgot about that one, could I get to look at the
> > patch again? I suppose it is ok, but it could potentially make the
> > driver refuse sampling from the input (returning -EIO) when there is no
> > signal there.
>
> well, i think you're referring to a different patch (yeah, i remember
> there was something like that, too).
>
> the patch below is the one i wanted to include.
> basically this changes only the name of clock source control element
> to be compliant to the definition in ControlNames.txt.

Ah, thanks. It looks perfectly ok to me, I suggest that you apply it.
Sorry for not reacting earlier...

/Anders


-------------------------------------------------------
This SF.Net email sponsored by: Free pre-built ASP.NET sites including
Data Reports, E-commerce, Portals, and Forums are available now.
Download today and enter to win an XBOX or Visual Studio .NET.
http://aspnet.click-url.com/go/psa00100006ave/direct;at.asp_061203_01/01