[PATCH] Jabra SPEAK 410 USB - no audio playback

[PATCH] Jabra SPEAK 410 USB - no audio playback

[Richard W.M. Jones]

[-- Attachment #1: Type: text/plain, Size: 1176 bytes --]


The attached patch fixes USB audio support for the Jabra SPEAK 410 USB.

The problem, as I understand it, is that the device contains a
Class-Specific Endpoint (CS_ENDPOINT) descriptor before the Endpoint
(ENDPOINT) descriptor.  The USB code all assumes that CS_ENDPOINT can
only appear after ENDPOINT.  Therefore the USB code divides up the
interface descriptor into "stuff before ENDPOINT" (in
interface->extra) and "stuff after ENDPOINT" (in
interface->endpoint[0]->extra).  For this device, this division does
not work.

You can see lsusb for my device here:
http://mailman.alsa-project.org/pipermail/alsa-devel/2011-December/047036.html
Notice the "** UNRECOGNIZED: 07 25 01 81 02 00 00" line which is the
CS_ENDPOINT descriptor.

The solution (which is a hack ...) is to also search interface->extra
looking for the missing descriptor.

For me, this fully enables the functions of this device.

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming blog: http://rwmj.wordpress.com
Fedora now supports 80 OCaml packages (the OPEN alternative to F#)
http://cocan.org/getting_started_with_ocaml_on_red_hat_and_fedora

[-- Attachment #2: linux-3.1.fc16-jabra.patch --]
[-- Type: text/plain, Size: 926 bytes --]

diff -uNrp kernel-3.1.fc16.orig/sound/usb/stream.c kernel-3.1.fc16.new/sound/usb/stream.c
--- kernel-3.1.fc16.orig/sound/usb/stream.c	2011-12-10 18:03:29.658729051 +0000
+++ kernel-3.1.fc16.new/sound/usb/stream.c	2011-12-10 18:08:41.468694907 +0000
@@ -164,6 +164,12 @@ static int parse_uac_endpoint_attributes
 
 	csep = snd_usb_find_desc(alts->endpoint[0].extra, alts->endpoint[0].extralen, NULL, USB_DT_CS_ENDPOINT);
 
+	/* Jabra SPEAK 410 USB has CS_ENDPOINT before ordinary ENDPOINT,
+	 * which confuses the USB descriptor parsing code.  Try looking
+	 * for CS_ENDPOINT in the interface->extra - RWMJ. */
+	if (!csep)
+		csep = snd_usb_find_desc(alts->extra, alts->extralen, NULL, USB_DT_CS_ENDPOINT);
+
 	/* Creamware Noah has this descriptor after the 2nd endpoint */
 	if (!csep && altsd->bNumEndpoints >= 2)
 		csep = snd_usb_find_desc(alts->endpoint[1].extra, alts->endpoint[1].extralen, NULL, USB_DT_CS_ENDPOINT);

Re: [alsa-devel] [PATCH] Jabra SPEAK 410 USB - no audio playback

[Richard W.M. Jones]

On Sat, Dec 10, 2011 at 07:45:39PM +0000, Richard W.M. Jones wrote:

> diff -uNrp kernel-3.1.fc16.orig/sound/usb/stream.c kernel-3.1.fc16.new/sound/usb/stream.c
> --- kernel-3.1.fc16.orig/sound/usb/stream.c	2011-12-10 18:03:29.658729051 +0000
> +++ kernel-3.1.fc16.new/sound/usb/stream.c	2011-12-10 18:08:41.468694907 +0000
> @@ -164,6 +164,12 @@ static int parse_uac_endpoint_attributes
>  
>  	csep = snd_usb_find_desc(alts->endpoint[0].extra, alts->endpoint[0].extralen, NULL, USB_DT_CS_ENDPOINT);
>  
> +	/* Jabra SPEAK 410 USB has CS_ENDPOINT before ordinary ENDPOINT,
> +	 * which confuses the USB descriptor parsing code.  Try looking
> +	 * for CS_ENDPOINT in the interface->extra - RWMJ. */
> +	if (!csep)
> +		csep = snd_usb_find_desc(alts->extra, alts->extralen, NULL, USB_DT_CS_ENDPOINT);
> +
>  	/* Creamware Noah has this descriptor after the 2nd endpoint */
>  	if (!csep && altsd->bNumEndpoints >= 2)
>  		csep = snd_usb_find_desc(alts->endpoint[1].extra, alts->endpoint[1].extralen, NULL, USB_DT_CS_ENDPOINT);


Anyone?  Tested on two different machines and it fixes the
audio output problem on both ...  If I've sent this to the
wrong list please let me know.

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming blog: http://rwmj.wordpress.com
Fedora now supports 80 OCaml packages (the OPEN alternative to F#)
http://cocan.org/getting_started_with_ocaml_on_red_hat_and_fedora

Re: [alsa-devel] [PATCH] Jabra SPEAK 410 USB - no audio playback

[Clemens Ladisch]

Richard W.M. Jones wrote:
> The attached patch fixes USB audio support for the Jabra SPEAK 410 USB.
>
> The problem, as I understand it, is that the device contains a
> Class-Specific Endpoint (CS_ENDPOINT) descriptor before the Endpoint
> (ENDPOINT) descriptor.  The USB code all assumes that CS_ENDPOINT can
> only appear after ENDPOINT.  Therefore the USB code divides up the
> interface descriptor into "stuff before ENDPOINT" (in
> interface->extra) and "stuff after ENDPOINT" (in
> interface->endpoint[0]->extra).  For this device, this division does
> not work.
>
> The solution (which is a hack ...) is to also search interface->extra
> looking for the missing descriptor.

The patch looks good.  Please run it through checkpatch.pl and provice
a Signed-off-by tag (see Documentation/SubmittingPatches).


Regards,
Clemens

Re: [alsa-devel] [PATCH] Jabra SPEAK 410 USB - no audio playback

[Richard W.M. Jones]

On Fri, Dec 16, 2011 at 02:49:05PM +0100, Clemens Ladisch wrote:
> Richard W.M. Jones wrote:
> > The attached patch fixes USB audio support for the Jabra SPEAK 410 USB.
> >
> > The problem, as I understand it, is that the device contains a
> > Class-Specific Endpoint (CS_ENDPOINT) descriptor before the Endpoint
> > (ENDPOINT) descriptor.  The USB code all assumes that CS_ENDPOINT can
> > only appear after ENDPOINT.  Therefore the USB code divides up the
> > interface descriptor into "stuff before ENDPOINT" (in
> > interface->extra) and "stuff after ENDPOINT" (in
> > interface->endpoint[0]->extra).  For this device, this division does
> > not work.
> >
> > The solution (which is a hack ...) is to also search interface->extra
> > looking for the missing descriptor.
> 
> The patch looks good.  Please run it through checkpatch.pl and provice
> a Signed-off-by tag (see Documentation/SubmittingPatches).

Sorry, I should have followed up on this list ...

Tom (in the CC line) opened a bug with Jabra, and they have issued a
firmware update.  Unfortunately this update requires some Windows
software to run.  I applied this, and it does appear to fix the audio
playback problem, though curiously it doesn't completely fix the
ENDPOINT/CS_ENDPOINT problem, so who knows what's really going on
there.

Anyway, should you want the update it is here:

https://bugzilla.redhat.com/show_bug.cgi?id=766714#c1

Thanks Tom!

Rich.

-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
virt-df lists disk usage of guests without needing to install any
software inside the virtual machine.  Supports Linux and Windows.
http://et.redhat.com/~rjones/virt-df/