[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87r113jal3.wl-tiwai@suse.de>
Date: Fri, 26 Aug 2022 17:20:08 +0200
From: Takashi Iwai <tiwai@...e.de>
To: Lennert Van Alboom <lennert@...alboom.org>
Cc: Jaroslav Kysela <perex@...ex.cz>, Takashi Iwai <tiwai@...e.com>,
"alsa-devel@...a-project.org" <alsa-devel@...a-project.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"ls.yoyo.m@...il.com" <ls.yoyo.m@...il.com>
Subject: Re: [PATCH] USB DAC broken since commit bf6313a0ff766925462e97b4e733d5952de02367 (5.10.0-rc5)
On Fri, 26 Aug 2022 11:07:55 +0200,
Lennert Van Alboom wrote:
>
> ------- Original Message -------
> On Friday, August 26th, 2022 at 08:05, Takashi Iwai <tiwai@...e.de> wrote:
> >
>
> > This option is passed per card instance, as the driver may hold
> > multiple cards. Check your /proc/asound/cards. The first argument of
> > quirk_flags is applied to the first USB-audio card, the second to the
> > second USB audio device, and so on.
> >
>
> > At best, give alsa-info.sh output before and after applying the
> > quirk. Run the script with --no-upload option and attach the
> > outputs.
> >
>
> > And, as a reference, you can see the existing quirk tables in
> > sound/usb/quirks.c.
> >
>
> >
>
> > Takashi
>
>
> Not sure why it failed before (didn't look further then, so possibly PEBKAC) but as you suggested, this works now (tested on debian's 5.19 kernel):
>
> # modprobe snd_usb_audio vid=0x2522 pid=0x0007 quirk_flags=0x10000
>
>
> Confirmed that setting a quirk_alias also works:
>
>
> # grep -A 1 "0x054c, 0x0b8c" linux/sound/usb/quirks.c
>
> DEVICE_FLG(0x054c, 0x0b8c, /* Sony WALKMAN NW-A45 DAC */
> QUIRK_FLAG_SET_IFACE_FIRST),
> # modprobe snd_usb_audio quirk_alias=25220007:054c0b8c
>
>
> I've added a patch for quirks.c to make this default behaviour for
> this device.
Good to hear. Will you submit the fix patch for the merge, or shall I
do it in my side?
> Looking at the Arch bug report at https://bugs.archlinux.org/task/70636, it's likely similar behaviour may be needed for pid=0009 (LH Labs Geek Pulse X 1V5) but I have no way of testing this. Added the reporter in CC.
It's likely OK to apply the same quirk, but it's certainly safer to
hear the test result, of course.
thanks,
Takashi
Powered by blists - more mailing lists