[PATCH 1/2] ALSA: snd_usb_caiaq: give better shortname

[PATCH 1/2] ALSA: snd_usb_caiaq: give better shortname

[Daniel Mack]

If not passed as module option, provide an own card ID with the newly
introduced snd_set_card_id() call.

This will prevent ALSA from calling choose_default_name() which only
takes the last part of a name containing whitespaces. This for example
caused 'Audio 4 DJ' to be shortened to 'DJ', which was not very
descriptive.

The implementation now takes the short name and removes all whitespaces
from it which is much nicer.

Signed-off-by: Daniel Mack <daniel@caiaq.de>
---
 sound/usb/caiaq/device.c |   21 +++++++++++++++++++--
 1 files changed, 19 insertions(+), 2 deletions(-)

diff --git a/sound/usb/caiaq/device.c b/sound/usb/caiaq/device.c
index 15052e4..43bd0db 100644
--- a/sound/usb/caiaq/device.c
+++ b/sound/usb/caiaq/device.c
@@ -382,10 +382,10 @@ static int create_card(struct usb_device* usb_dev, struct snd_card **cardp)
 
 static int __devinit init_card(struct snd_usb_caiaqdev *dev)
 {
-	char usbpath[32];
+	char *c, usbpath[32];
 	struct usb_device *usb_dev = dev->chip.dev;
 	struct snd_card *card = dev->chip.card;
-	int err;
+	int err, len;
 
 	if (usb_set_interface(usb_dev, 0, 1) != 0) {
 		log("can't set alt interface.\n");
@@ -427,6 +427,23 @@ static int __devinit init_card(struct snd_usb_caiaqdev *dev)
 	strlcpy(card->driver, MODNAME, sizeof(card->driver));
 	strlcpy(card->shortname, dev->product_name, sizeof(card->shortname));
 
+	/* if the id was not passed as module option, fill it with a shortened
+	 * version of the product string which does not contain any
+	 * whitespaces */
+
+	if (*card->id == '\0') {
+		char id[sizeof(card->id)];
+
+		memset(id, 0, sizeof(id));
+
+		for (c = card->shortname, len = 0;
+			*c && len < sizeof(card->id); c++)
+			if (*c != ' ')
+				id[len++] = *c;
+
+		snd_card_set_id(card, id);
+	}
+
 	usb_make_path(usb_dev, usbpath, sizeof(usbpath));
 	snprintf(card->longname, sizeof(card->longname),
 		       "%s %s (%s)",
-- 
1.6.3.1

[PATCH 2/2] ALSA: snd_usb_caiaq: bump version number

[Daniel Mack]

Signed-off-by: Daniel Mack <daniel@caiaq.de>
---
 sound/usb/caiaq/device.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/sound/usb/caiaq/device.c b/sound/usb/caiaq/device.c
index 43bd0db..b9a2b31 100644
--- a/sound/usb/caiaq/device.c
+++ b/sound/usb/caiaq/device.c
@@ -35,7 +35,7 @@
 #include "input.h"
 
 MODULE_AUTHOR("Daniel Mack <daniel@caiaq.de>");
-MODULE_DESCRIPTION("caiaq USB audio, version 1.3.14");
+MODULE_DESCRIPTION("caiaq USB audio, version 1.3.15");
 MODULE_LICENSE("GPL");
 MODULE_SUPPORTED_DEVICE("{{Native Instruments, RigKontrol2},"
 			 "{Native Instruments, RigKontrol3},"
-- 
1.6.3.1

Re: [PATCH 1/2] ALSA: snd_usb_caiaq: give better shortname

[Takashi Iwai]

At Tue,  2 Jun 2009 12:36:39 +0200,
Daniel Mack wrote:
> 
> If not passed as module option, provide an own card ID with the newly
> introduced snd_set_card_id() call.
> 
> This will prevent ALSA from calling choose_default_name() which only
> takes the last part of a name containing whitespaces. This for example
> caused 'Audio 4 DJ' to be shortened to 'DJ', which was not very
> descriptive.
> 
> The implementation now takes the short name and removes all whitespaces
> from it which is much nicer.
> 
> Signed-off-by: Daniel Mack <daniel@caiaq.de>
> ---
>  sound/usb/caiaq/device.c |   21 +++++++++++++++++++--
>  1 files changed, 19 insertions(+), 2 deletions(-)
> 
> diff --git a/sound/usb/caiaq/device.c b/sound/usb/caiaq/device.c
> index 15052e4..43bd0db 100644
> --- a/sound/usb/caiaq/device.c
> +++ b/sound/usb/caiaq/device.c
> @@ -382,10 +382,10 @@ static int create_card(struct usb_device* usb_dev, struct snd_card **cardp)
>  
>  static int __devinit init_card(struct snd_usb_caiaqdev *dev)
>  {
> -	char usbpath[32];
> +	char *c, usbpath[32];
>  	struct usb_device *usb_dev = dev->chip.dev;
>  	struct snd_card *card = dev->chip.card;
> -	int err;
> +	int err, len;
>  
>  	if (usb_set_interface(usb_dev, 0, 1) != 0) {
>  		log("can't set alt interface.\n");
> @@ -427,6 +427,23 @@ static int __devinit init_card(struct snd_usb_caiaqdev *dev)
>  	strlcpy(card->driver, MODNAME, sizeof(card->driver));
>  	strlcpy(card->shortname, dev->product_name, sizeof(card->shortname));
>  
> +	/* if the id was not passed as module option, fill it with a shortened
> +	 * version of the product string which does not contain any
> +	 * whitespaces */
> +
> +	if (*card->id == '\0') {

Hm, at this place, card->id should be never an empty string so
this is always true.


> +		char id[sizeof(card->id)];
> +
> +		memset(id, 0, sizeof(id));
> +
> +		for (c = card->shortname, len = 0;
> +			*c && len < sizeof(card->id); c++)
> +			if (*c != ' ')
> +				id[len++] = *c;

... and here you change the id string unconditionally no matter whether 
it's given explicitly or not.  It should check id[devnum] instead of
card->id.


thanks,

Takashi

Re: [PATCH 1/2] ALSA: snd_usb_caiaq: give better shortname

[Daniel Mack]

On Tue, Jun 02, 2009 at 01:01:35PM +0200, Takashi Iwai wrote:
> > +	/* if the id was not passed as module option, fill it with a shortened
> > +	 * version of the product string which does not contain any
> > +	 * whitespaces */
> > +
> > +	if (*card->id == '\0') {
> 
> Hm, at this place, card->id should be never an empty string so
> this is always true.

Are you sure about that? The same condition is used in
snd_card_register() which is called _after_ the code in question from my
driver.

And the id should not be zero if snd_card_create() was called with an id
argument given. So the condition above should be false if the id was passed
before. Or do I miss anything?

Daniel

Re: [PATCH 1/2] ALSA: snd_usb_caiaq: give better shortname

[Takashi Iwai]

At Tue, 2 Jun 2009 13:55:59 +0200,
Daniel Mack wrote:
> 
> On Tue, Jun 02, 2009 at 01:01:35PM +0200, Takashi Iwai wrote:
> > > +	/* if the id was not passed as module option, fill it with a shortened
> > > +	 * version of the product string which does not contain any
> > > +	 * whitespaces */
> > > +
> > > +	if (*card->id == '\0') {
> > 
> > Hm, at this place, card->id should be never an empty string so
> > this is always true.
> 
> Are you sure about that? The same condition is used in
> snd_card_register() which is called _after_ the code in question from my
> driver.

Ah, no I was wrong.  Actually it's set in snd_card_register()
so your method would work.

Now applied the patches.


thanks,

Takashi