au88x0 Init before create components

au88x0 Init before create components

[Alan Horstmann]

[-- Attachment #1: Type: text/plain, Size: 523 bytes --]

Patch as suggested by Raymond in one of his copious bug-track notes.

Change the order in vortex_probe to init the card details before creating the 
components, meaning for example that card->shortname is available when 
registering the midi port.

I have also added extra to card->shortname, and a line to overwrite the midi 
name following snd_mpu401_uart_new, for consideration.

Do I have the correct address to cc mjander?

au88x0-init-before-comps-et-al.patch

Signed-off-by: Alan Horstmann <gineera@aspect135.co.uk>

[-- Attachment #2: au88x0-init-before-comps-et-al.patch --]
[-- Type: text/x-diff, Size: 2488 bytes --]

diff -ru alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c
--- alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c	2006-01-13 17:16:29.000000000 +0000
+++ alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c	2006-04-24 08:12:16.000000000 +0100
@@ -261,7 +261,28 @@
 		return err;
 	}
 	snd_vortex_workaround(pci, pcifix[dev]);
-	// (4) Alloc components.
+	// (4)
+	strcpy(card->driver, CARD_NAME_SHORT);
+	sprintf(card->shortname, "Aureal Vortex %s", CARD_NAME_SHORT);
+	sprintf(card->longname, "%s at 0x%lx irq %i",
+		card->shortname, chip->io, chip->irq);
+
+	if ((err = pci_read_config_word(pci, PCI_DEVICE_ID,
+				  &(chip->device))) < 0) {
+		snd_card_free(card);
+		return err;
+	}
+	if ((err = pci_read_config_word(pci, PCI_VENDOR_ID,
+				  &(chip->vendor))) < 0) {
+		snd_card_free(card);
+		return err;
+	}
+	if ((err = pci_read_config_byte(pci, PCI_REVISION_ID,
+				  &(chip->rev))) < 0) {
+		snd_card_free(card);
+		return err;
+	}
+	// (5) Alloc components.
 	// ADB pcm.
 	if ((err = snd_vortex_new_pcm(chip, VORTEX_PCM_ADB, NR_ADB)) < 0) {
 		snd_card_free(card);
@@ -322,27 +343,6 @@
 	}
 #endif
 
-	// (5)
-	strcpy(card->driver, CARD_NAME_SHORT);
-	strcpy(card->shortname, CARD_NAME_SHORT);
-	sprintf(card->longname, "%s at 0x%lx irq %i",
-		card->shortname, chip->io, chip->irq);
-
-	if ((err = pci_read_config_word(pci, PCI_DEVICE_ID,
-				  &(chip->device))) < 0) {
-		snd_card_free(card);
-		return err;
-	}	
-	if ((err = pci_read_config_word(pci, PCI_VENDOR_ID,
-				  &(chip->vendor))) < 0) {
-		snd_card_free(card);
-		return err;
-	}
-	if ((err = pci_read_config_byte(pci, PCI_REVISION_ID,
-				  &(chip->rev))) < 0) {
-		snd_card_free(card);
-		return err;
-	}
 #ifdef CHIP_AU8830
 	if ((chip->rev) != 0xfe && (chip->rev) != 0xfa) {
 		printk(KERN_ALERT
diff -ru alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c
--- alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c	2006-01-13 17:16:29.000000000 +0000
+++ alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c	2006-04-24 07:57:51.000000000 +0100
@@ -107,6 +107,9 @@
 	mpu = rmidi->private_data;
 	mpu->cport = (unsigned long)(vortex->mmio + VORTEX_MIDI_CMD);
 #endif
+	/* Overwrite MIDI name */
+	snprintf(rmidi->name, sizeof(rmidi->name), "MIDI %s %d", CARD_NAME_SHORT , vortex->card->number);
+
 	vortex->rmidi = rmidi;
 	return 0;
 }

Re: au88x0 Init before create components

[Takashi Iwai]

At Wed, 26 Apr 2006 14:14:52 +0100,
Alan Horstmann wrote:
> 
> Patch as suggested by Raymond in one of his copious bug-track notes.
> 
> Change the order in vortex_probe to init the card details before creating the 
> components, meaning for example that card->shortname is available when 
> registering the midi port.

The detection of vendor, device and rev is irrelevant with the
definition of driver, shortname and longname.

Looking at the code a bit carefully, we can find that chip->vendor and
chip->device are not really used in the code.  The only referring part
is IS_BAD_CHIP() macro in au88x0.h, and it's used nowhere.  Moreover,
the values are identical with pci->device and pci->vendor.
So, basically we can remove these fields.  But it's another story.

Anyway, I'd suggest to move the definition of driver, shortname and
longname there with a short comment that it's needed for mpu401
instance, and keep the rest as it is.

Also, I feel "Aureal au8810 MIDI 0" is more intuitive than
"MIDI Aureal au8810 0".  It's a matter of taste, though...


Thanks,

Takashi



> 
> I have also added extra to card->shortname, and a line to overwrite the midi 
> name following snd_mpu401_uart_new, for consideration.
> 
> Do I have the correct address to cc mjander?
> 
> au88x0-init-before-comps-et-al.patch
> 
> Signed-off-by: Alan Horstmann <gineera@aspect135.co.uk>
> [2 au88x0-init-before-comps-et-al.patch <text/x-diff; us-ascii (7bit)>]
> diff -ru alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c
> --- alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c	2006-01-13 17:16:29.000000000 +0000
> +++ alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c	2006-04-24 08:12:16.000000000 +0100
> @@ -261,7 +261,28 @@
>  		return err;
>  	}
>  	snd_vortex_workaround(pci, pcifix[dev]);
> -	// (4) Alloc components.
> +	// (4)
> +	strcpy(card->driver, CARD_NAME_SHORT);
> +	sprintf(card->shortname, "Aureal Vortex %s", CARD_NAME_SHORT);
> +	sprintf(card->longname, "%s at 0x%lx irq %i",
> +		card->shortname, chip->io, chip->irq);
> +
> +	if ((err = pci_read_config_word(pci, PCI_DEVICE_ID,
> +				  &(chip->device))) < 0) {
> +		snd_card_free(card);
> +		return err;
> +	}
> +	if ((err = pci_read_config_word(pci, PCI_VENDOR_ID,
> +				  &(chip->vendor))) < 0) {
> +		snd_card_free(card);
> +		return err;
> +	}
> +	if ((err = pci_read_config_byte(pci, PCI_REVISION_ID,
> +				  &(chip->rev))) < 0) {
> +		snd_card_free(card);
> +		return err;
> +	}
> +	// (5) Alloc components.
>  	// ADB pcm.
>  	if ((err = snd_vortex_new_pcm(chip, VORTEX_PCM_ADB, NR_ADB)) < 0) {
>  		snd_card_free(card);
> @@ -322,27 +343,6 @@
>  	}
>  #endif
>  
> -	// (5)
> -	strcpy(card->driver, CARD_NAME_SHORT);
> -	strcpy(card->shortname, CARD_NAME_SHORT);
> -	sprintf(card->longname, "%s at 0x%lx irq %i",
> -		card->shortname, chip->io, chip->irq);
> -
> -	if ((err = pci_read_config_word(pci, PCI_DEVICE_ID,
> -				  &(chip->device))) < 0) {
> -		snd_card_free(card);
> -		return err;
> -	}	
> -	if ((err = pci_read_config_word(pci, PCI_VENDOR_ID,
> -				  &(chip->vendor))) < 0) {
> -		snd_card_free(card);
> -		return err;
> -	}
> -	if ((err = pci_read_config_byte(pci, PCI_REVISION_ID,
> -				  &(chip->rev))) < 0) {
> -		snd_card_free(card);
> -		return err;
> -	}
>  #ifdef CHIP_AU8830
>  	if ((chip->rev) != 0xfe && (chip->rev) != 0xfa) {
>  		printk(KERN_ALERT
> diff -ru alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c
> --- alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c	2006-01-13 17:16:29.000000000 +0000
> +++ alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c	2006-04-24 07:57:51.000000000 +0100
> @@ -107,6 +107,9 @@
>  	mpu = rmidi->private_data;
>  	mpu->cport = (unsigned long)(vortex->mmio + VORTEX_MIDI_CMD);
>  #endif
> +	/* Overwrite MIDI name */
> +	snprintf(rmidi->name, sizeof(rmidi->name), "MIDI %s %d", CARD_NAME_SHORT , vortex->card->number);
> +
>  	vortex->rmidi = rmidi;
>  	return 0;
>  }


-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: au88x0 Init before create components

[Manuel Jander]

Hi,

This is my current and hopefuly "for ever" email address.
(hmm, hopefuly no spammer gets this one: _manuel_._jander_|@_gmail|.|
com) (remove _ and | characters).
My address at embedded.cl no longer exists. A am very pleased about that
someone is posting patches to the au88x0 drivers :)

By the way I'm living in Erlangen now, far close to the ALSA guys in
Nürnberg.

Best Regards,
Manuel Jander

On Wed, 2006-04-26 at 14:14 +0100, Alan Horstmann wrote:
> Patch as suggested by Raymond in one of his copious bug-track notes.
> 
> Change the order in vortex_probe to init the card details before creating the 
> components, meaning for example that card->shortname is available when 
> registering the midi port.
> 
> I have also added extra to card->shortname, and a line to overwrite the midi 
> name following snd_mpu401_uart_new, for consideration.
> 
> Do I have the correct address to cc mjander?
> 
> au88x0-init-before-comps-et-al.patch
> 
> Signed-off-by: Alan Horstmann <gineera@aspect135.co.uk>



-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: au88x0 Init before create components

[Alan Horstmann]

> At Wed, 26 Apr 2006 14:14:52 +0100,
>
> Alan Horstmann wrote:
> > I have also added extra to card->shortname, and a line to overwrite the
> > midi name following snd_mpu401_uart_new, for consideration.

Raymond wrote:

> Refer to au88x0-init-before-comps-et-al.patch,
> You should not change the card->shortname
> THis shortname is used to access the /usr/share/alsa/conf/AU88X0.conf

Or is it card->driver that accesses .conf files?  Other cards operate with 
shortnames different from .conf file name.  (And of course it works fine in 
my tests)

Alan



-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: Re: au88x0 Init before create components

[Takashi Iwai]

At Wed, 26 Apr 2006 21:00:58 +0100,
Alan Horstmann wrote:
> 
> > At Wed, 26 Apr 2006 14:14:52 +0100,
> >
> > Alan Horstmann wrote:
> > > I have also added extra to card->shortname, and a line to overwrite the
> > > midi name following snd_mpu401_uart_new, for consideration.
> 
> Raymond wrote:
> 
> > Refer to au88x0-init-before-comps-et-al.patch,
> > You should not change the card->shortname
> > THis shortname is used to access the /usr/share/alsa/conf/AU88X0.conf
> 
> Or is it card->driver that accesses .conf files?  Other cards operate with 
> shortnames different from .conf file name.  (And of course it works fine in 
> my tests)

Yes, it is card->driver.  shortname isn't referred from the config
file.  The change of shortname is safe (so far).


Takashi


-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: au88x0 Init before create components

[Alan Horstmann]

[-- Attachment #1: Type: text/plain, Size: 776 bytes --]

On Wednesday 26 April 2006 16:05, you wrote:
>
> Anyway, I'd suggest to move the definition of driver, shortname and
> longname there with a short comment that it's needed for mpu401
> instance, and keep the rest as it is.
>
> Also, I feel "Aureal au8810 MIDI 0" is more intuitive than
> "MIDI Aureal au8810 0".  It's a matter of taste, though...

Reduced patch is attached

Regards  Alan

Change the order in vortex_probe to set the card details before creating the 
components, meaning for example that card->shortname is available when 
registering the midi port.

I have also added extra to card->shortname, and a line to overwrite the midi 
name following snd_mpu401_uart_new.

au88x0-card-details-midi-name.patch

Signed-off-by: Alan Horstmann <gineera@aspect135.co.uk>

[-- Attachment #2: au88x0-card-details-midi-name.patch --]
[-- Type: text/x-diff, Size: 1708 bytes --]

diff -ru alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c
--- alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c	2006-01-13 17:16:29.000000000 +0000
+++ alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0.c	2006-04-24 08:12:16.000000000 +0100
@@ -261,6 +261,13 @@
 		return err;
 	}
 	snd_vortex_workaround(pci, pcifix[dev]);
+
+	// Card details needed in snd_vortex_midi
+	strcpy(card->driver, CARD_NAME_SHORT);
+	sprintf(card->shortname, "Aureal Vortex %s", CARD_NAME_SHORT);
+	sprintf(card->longname, "%s at 0x%lx irq %i",
+		card->shortname, chip->io, chip->irq);
+
	// (4) Alloc components.
 	// ADB pcm.
 	if ((err = snd_vortex_new_pcm(chip, VORTEX_PCM_ADB, NR_ADB)) < 0) {
@@ -323,11 +330,6 @@
 #endif

	// (5)
-	strcpy(card->driver, CARD_NAME_SHORT);
-	strcpy(card->shortname, CARD_NAME_SHORT);
-	sprintf(card->longname, "%s at 0x%lx irq %i",
-		card->shortname, chip->io, chip->irq);
-
	if ((err = pci_read_config_word(pci, PCI_DEVICE_ID,
				  &(chip->device))) < 0) {
		snd_card_free(card);
diff -ru alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c
--- alsa-driver-orig-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c	2006-01-13 17:16:29.000000000 +0000
+++ alsa-driver-1.0.11rc3/alsa-kernel/pci/au88x0/au88x0_mpu401.c	2006-04-24 07:57:51.000000000 +0100
@@ -107,6 +107,9 @@
 	mpu = rmidi->private_data;
 	mpu->cport = (unsigned long)(vortex->mmio + VORTEX_MIDI_CMD);
 #endif
+	/* Overwrite MIDI name */
+	snprintf(rmidi->name, sizeof(rmidi->name), "%s MIDI %d", CARD_NAME_SHORT , vortex->card->number);
+
 	vortex->rmidi = rmidi;
 	return 0;
 }

Re: Re: au88x0 Init before create components

[Takashi Iwai]

At Mon, 1 May 2006 22:17:48 +0100,
Alan Horstmann wrote:
> 
> [1  <text/plain; iso-8859-1 (7bit)>]
> On Wednesday 26 April 2006 16:05, you wrote:
> >
> > Anyway, I'd suggest to move the definition of driver, shortname and
> > longname there with a short comment that it's needed for mpu401
> > instance, and keep the rest as it is.
> >
> > Also, I feel "Aureal au8810 MIDI 0" is more intuitive than
> > "MIDI Aureal au8810 0".  It's a matter of taste, though...
> 
> Reduced patch is attached

Thanks.

I think it's better to remove the card number from the shortname
field.  You can add it at copying to rmidi->name, instead.

If you have no objection, I'll change and commit it.


Takashi


-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: Re: au88x0 Init before create components

[Alan Horstmann]

On Wednesday 03 May 2006 13:11, you wrote:
> At Mon, 1 May 2006 22:17:48 +0100,
>
> Alan Horstmann wrote:
> >
> > Reduced patch is attached
>
> I think it's better to remove the card number from the shortname
> field.  You can add it at copying to rmidi->name, instead.

That is what the patch does -card number isn't added to shortname, but just 
rmidi->name.

Alan



-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: Re: au88x0 Init before create components

[Takashi Iwai]

At Wed, 3 May 2006 15:55:57 +0100,
Alan Horstmann wrote:
> 
> On Wednesday 03 May 2006 13:11, you wrote:
> > At Mon, 1 May 2006 22:17:48 +0100,
> >
> > Alan Horstmann wrote:
> > >
> > > Reduced patch is attached
> >
> > I think it's better to remove the card number from the shortname
> > field.  You can add it at copying to rmidi->name, instead.
> 
> That is what the patch does -card number isn't added to shortname, but just 
> rmidi->name.

Ah OK, then it's fine.


thanks,

Takashi


-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642