Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Bug#1027483: linux-image-5.10.0-20-amd64: sound with firmware-sof-intel on Intel Lake stops working with newer kernel
117 views
Skip to first unread message
Malvin Gattinger
Jan 1, 2023, 8:50:03 AM1/1/23
to
Package: linux-image-5.10.0-20-amd64
Version: 5.10.158-2
Severity: normal
X-Debbugs-Cc: mal...@w4eg.eu
Dear Maintainer,
I am unsure if this problem belongs to linux-image-amd64 as a regression or is actually a but in firmware-sof-intel.
On an HP ProBook with this sound card (from lspci):
00:1f.3 Multimedia audio controller: Intel Corporation Tiger Lake-LP Smart Sound Technology Audio Controller (rev 20)
Updating from linux-image-5.10.0-19-amd64 to linux-image-5.10.0-20-amd64 breaks sound.
Sound output from VLC works fine immediately after boot, but using the microphone or other audio use cases later disables all output. Unfortunately the only reproducible example on this laptop was to open the proprietary zoom video-conferencing software which immediately stops all output, but other symptoms are discussed here:
https://forums.debian.net/viewtopic.php?t=153585
https://forums.debian.net/viewtopic.php?p=764594
PulseAudio claims that the output is still running, i.e. pavucontrol and KDE audio settings still show a moving bar for the output.
The only workaround for now is to use the older kernel from linux-image-5.10.0-19-amd64.
kind regards and happy new year,
Malvin Gattinger
-- System Information:
Debian Release: 11.6
APT prefers stable-updates
APT policy: (500, 'stable-updates'), (500, 'stable-security'), (500, 'stable')
Architecture: amd64 (x86_64)
Foreign Architectures: i386
Kernel: Linux 5.10.0-19-amd64 (SMP w/8 CPU threads)
Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled
Versions of packages linux-image-5.10.0-20-amd64 depends on:
ii initramfs-tools [linux-initramfs-tool] 0.140
ii kmod 28-1
ii linux-base 4.6
Versions of packages linux-image-5.10.0-20-amd64 recommends:
ii apparmor 2.13.6-10
ii firmware-linux-free 20200122-1
Versions of packages linux-image-5.10.0-20-amd64 suggests:
pn debian-kernel-handbook <none>
ii grub-efi-amd64 2.06-3~deb11u5
pn linux-doc-5.10 <none>
Version: 5.10.158-2
Severity: normal
X-Debbugs-Cc: mal...@w4eg.eu
Dear Maintainer,
I am unsure if this problem belongs to linux-image-amd64 as a regression or is actually a but in firmware-sof-intel.
On an HP ProBook with this sound card (from lspci):
00:1f.3 Multimedia audio controller: Intel Corporation Tiger Lake-LP Smart Sound Technology Audio Controller (rev 20)
Updating from linux-image-5.10.0-19-amd64 to linux-image-5.10.0-20-amd64 breaks sound.
Sound output from VLC works fine immediately after boot, but using the microphone or other audio use cases later disables all output. Unfortunately the only reproducible example on this laptop was to open the proprietary zoom video-conferencing software which immediately stops all output, but other symptoms are discussed here:
https://forums.debian.net/viewtopic.php?t=153585
https://forums.debian.net/viewtopic.php?p=764594
PulseAudio claims that the output is still running, i.e. pavucontrol and KDE audio settings still show a moving bar for the output.
The only workaround for now is to use the older kernel from linux-image-5.10.0-19-amd64.
kind regards and happy new year,
Malvin Gattinger
-- System Information:
Debian Release: 11.6
APT prefers stable-updates
APT policy: (500, 'stable-updates'), (500, 'stable-security'), (500, 'stable')
Architecture: amd64 (x86_64)
Foreign Architectures: i386
Kernel: Linux 5.10.0-19-amd64 (SMP w/8 CPU threads)
Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled
Versions of packages linux-image-5.10.0-20-amd64 depends on:
ii initramfs-tools [linux-initramfs-tool] 0.140
ii kmod 28-1
ii linux-base 4.6
Versions of packages linux-image-5.10.0-20-amd64 recommends:
ii apparmor 2.13.6-10
ii firmware-linux-free 20200122-1
Versions of packages linux-image-5.10.0-20-amd64 suggests:
pn debian-kernel-handbook <none>
ii grub-efi-amd64 2.06-3~deb11u5
pn linux-doc-5.10 <none>
Salvatore Bonaccorso
Jan 1, 2023, 10:40:03 AM1/1/23
to
Control: reassign -1 src:linux 5.10.158-2
Hi Malvin,
On Sun, Jan 01, 2023 at 02:36:28PM +0100, Malvin Gattinger wrote:
> Package: linux-image-5.10.0-20-amd64
> Version: 5.10.158-2
> Severity: normal
> X-Debbugs-Cc: mal...@w4eg.eu
>
> Dear Maintainer,
>
> I am unsure if this problem belongs to linux-image-amd64 as a
> regression or is actually a but in firmware-sof-intel.
>
> On an HP ProBook with this sound card (from lspci):
>
> 00:1f.3 Multimedia audio controller: Intel Corporation Tiger Lake-LP
> Smart Sound Technology Audio Controller (rev 20)
>
> Updating from linux-image-5.10.0-19-amd64 to
> linux-image-5.10.0-20-amd64 breaks sound.
>
> Sound output from VLC works fine immediately after boot, but using
> the microphone or other audio use cases later disables all output.
> Unfortunately the only reproducible example on this laptop was to
> open the proprietary zoom video-conferencing software which
> immediately stops all output, but other symptoms are discussed here:
>
> https://forums.debian.net/viewtopic.php?t=153585
>
> https://forums.debian.net/viewtopic.php?p=764594
>
> PulseAudio claims that the output is still running, i.e. pavucontrol
> and KDE audio settings still show a moving bar for the output.
>
> The only workaround for now is to use the older kernel from
> linux-image-5.10.0-19-amd64.
To pin point the possibly breaking change in the kernel, would you be
able to build kernels, first both upstream kernels mathcing the
5.10.149 (for the 19 ABI one), and 5.10.158 and then bisect betweeen
the two?
https://wiki.debian.org/DebianKernel/GitBisect
Regards,
Salvatore
Hi Malvin,
On Sun, Jan 01, 2023 at 02:36:28PM +0100, Malvin Gattinger wrote:
> Package: linux-image-5.10.0-20-amd64
> Version: 5.10.158-2
> Severity: normal
> X-Debbugs-Cc: mal...@w4eg.eu
>
> Dear Maintainer,
>
> I am unsure if this problem belongs to linux-image-amd64 as a
> regression or is actually a but in firmware-sof-intel.
>
> On an HP ProBook with this sound card (from lspci):
>
> 00:1f.3 Multimedia audio controller: Intel Corporation Tiger Lake-LP
> Smart Sound Technology Audio Controller (rev 20)
>
> Updating from linux-image-5.10.0-19-amd64 to
> linux-image-5.10.0-20-amd64 breaks sound.
>
> Sound output from VLC works fine immediately after boot, but using
> the microphone or other audio use cases later disables all output.
> Unfortunately the only reproducible example on this laptop was to
> open the proprietary zoom video-conferencing software which
> immediately stops all output, but other symptoms are discussed here:
>
> https://forums.debian.net/viewtopic.php?t=153585
>
> https://forums.debian.net/viewtopic.php?p=764594
>
> PulseAudio claims that the output is still running, i.e. pavucontrol
> and KDE audio settings still show a moving bar for the output.
>
> The only workaround for now is to use the older kernel from
> linux-image-5.10.0-19-amd64.
able to build kernels, first both upstream kernels mathcing the
5.10.149 (for the 19 ABI one), and 5.10.158 and then bisect betweeen
the two?
https://wiki.debian.org/DebianKernel/GitBisect
Regards,
Salvatore
Salvatore Bonaccorso
Jan 1, 2023, 10:50:04 AM1/1/23
to
Control: tags -1 + moreinfo
Following that, I wonder if #1027483, #1026834 (and maybe #1010733)
will have all the same underlying cause.
Regards,
Salvatore
will have all the same underlying cause.
Regards,
Salvatore
Malvin Gattinger
Jan 1, 2023, 1:40:04 PM1/1/23
to
Dear Salvatore,
> https://wiki.debian.org/DebianKernel/GitBisect
Thank you for the quick reply and suggestion to bisect this. But
(un)fortunately the laptop where the problem occurs is not my own
and I have neither the experience nor time to compile and try
kernels on it. Sorry if this would be my job, now that I made a
report. I only saw those forum posts so far and was mostly hoping
to get other eyes on this by filing a bug.
Two more things I can add:
- without "firmware-sof-intel" no audio devices at all are shown
by pulseaudio.
- with the 6.0 kernel from bullseye-backports there are also no
audio devices at all, even when "firmware-sof-intel" is
installed.
kind regards,
Malvin
> https://wiki.debian.org/DebianKernel/GitBisect
Thank you for the quick reply and suggestion to bisect this. But
(un)fortunately the laptop where the problem occurs is not my own
and I have neither the experience nor time to compile and try
kernels on it. Sorry if this would be my job, now that I made a
report. I only saw those forum posts so far and was mostly hoping
to get other eyes on this by filing a bug.
Two more things I can add:
- without "firmware-sof-intel" no audio devices at all are shown
by pulseaudio.
- with the 6.0 kernel from bullseye-backports there are also no
audio devices at all, even when "firmware-sof-intel" is
installed.
kind regards,
Malvin
PÁLFFY Dániel
Jan 1, 2023, 3:50:04 PM1/1/23
to
$ git bisect view
commit c34db0d6b88b1da95e7ab3353e674f4f574cccee (HEAD, refs/bisect/bad)
Author: Richard Fitzgerald <r...@opensource.cirrus.com>
Date: Fri Nov 4 13:22:13 2022 +0000
ASoC: soc-pcm: Don't zero TDM masks in __soc_pcm_open()
[ Upstream commit 39bd801d6908900e9ab0cdc2655150f95ddd4f1a ]
The DAI tx_mask and rx_mask are set by snd_soc_dai_set_tdm_slot()
and used by later code that depends on the TDM settings. So
__soc_pcm_open() should not be obliterating those mask values.
The code in __soc_pcm_hw_params() uses these masks to calculate the
active channels so that only the AIF_IN/AIF_OUT widgets for the
active TDM slots are enabled. The zeroing of the masks in
__soc_pcm_open() disables this functionality so all AIF widgets
were enabled even for channels that are not assigned to a TDM slot.
Signed-off-by: Richard Fitzgerald <r...@opensource.cirrus.com>
Fixes: 2e5894d73789 ("ASoC: pcm: Add support for DAI multicodec")
Link: https://lore.kernel.org/r/2022110413221...@opensource.cirrus.com
Signed-off-by: Mark Brown <bro...@kernel.org>
Signed-off-by: Sasha Levin <sas...@kernel.org>
commit c34db0d6b88b1da95e7ab3353e674f4f574cccee (HEAD, refs/bisect/bad)
Author: Richard Fitzgerald <r...@opensource.cirrus.com>
Date: Fri Nov 4 13:22:13 2022 +0000
ASoC: soc-pcm: Don't zero TDM masks in __soc_pcm_open()
[ Upstream commit 39bd801d6908900e9ab0cdc2655150f95ddd4f1a ]
The DAI tx_mask and rx_mask are set by snd_soc_dai_set_tdm_slot()
and used by later code that depends on the TDM settings. So
__soc_pcm_open() should not be obliterating those mask values.
The code in __soc_pcm_hw_params() uses these masks to calculate the
active channels so that only the AIF_IN/AIF_OUT widgets for the
active TDM slots are enabled. The zeroing of the masks in
__soc_pcm_open() disables this functionality so all AIF widgets
were enabled even for channels that are not assigned to a TDM slot.
Signed-off-by: Richard Fitzgerald <r...@opensource.cirrus.com>
Fixes: 2e5894d73789 ("ASoC: pcm: Add support for DAI multicodec")
Link: https://lore.kernel.org/r/2022110413221...@opensource.cirrus.com
Signed-off-by: Mark Brown <bro...@kernel.org>
Signed-off-by: Sasha Levin <sas...@kernel.org>
PÁLFFY Dániel
Jan 1, 2023, 5:30:04 PM1/1/23
to
Kernel bug: https://bugzilla.kernel.org/show_bug.cgi?id=216861
6.1.x seems to work for me, despite including the same commit.
Apparently a similar set of machines were affected by the different https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1022544 in 6.0.x including linux-image-6.0.0-0.deb11.2-amd64 still in bullseye-backports.
Salvatore Bonaccorso
Jan 2, 2023, 5:40:03 AM1/2/23
to
Hi Malvin,
On Sun, Jan 01, 2023 at 07:27:31PM +0100, Malvin Gattinger wrote:
> Dear Salvatore,
>
> > https://wiki.debian.org/DebianKernel/GitBisect
>
> Thank you for the quick reply and suggestion to bisect this. But
> (un)fortunately the laptop where the problem occurs is not my own and I have
> neither the experience nor time to compile and try kernels on it. Sorry if
> this would be my job, now that I made a report. I only saw those forum posts
> so far and was mostly hoping to get other eyes on this by filing a bug.
No worries, was more the question if you can further help pinpointing
the issue. Dániel bisected the issue down to a change introduced in
5.10.157.
Regards,
Salvatore
On Sun, Jan 01, 2023 at 07:27:31PM +0100, Malvin Gattinger wrote:
> Dear Salvatore,
>
> > https://wiki.debian.org/DebianKernel/GitBisect
>
> Thank you for the quick reply and suggestion to bisect this. But
> (un)fortunately the laptop where the problem occurs is not my own and I have
> neither the experience nor time to compile and try kernels on it. Sorry if
> this would be my job, now that I made a report. I only saw those forum posts
> so far and was mostly hoping to get other eyes on this by filing a bug.
the issue. Dániel bisected the issue down to a change introduced in
5.10.157.
Regards,
Salvatore
Tomas Pospisek
Jan 25, 2023, 5:20:04 AM1/25/23
to
and I guess #1027483 can be closed as it is still open along with #1027430
?
On Wed, 25 Jan 2023, Tomas Pospisek wrote:
> Woah guys, a new Debian stable kernel came my way and finally sound works
> again consistently. Many, many, many thanks to you for bisecting, patching
> and shipping the fix. Muito, muito obrigado!
> *t
>
?
On Wed, 25 Jan 2023, Tomas Pospisek wrote:
> Woah guys, a new Debian stable kernel came my way and finally sound works
> again consistently. Many, many, many thanks to you for bisecting, patching
> and shipping the fix. Muito, muito obrigado!
> *t
>
Tomas Pospisek
Jan 25, 2023, 5:20:05 AM1/25/23
to
Salvatore Bonaccorso
Jan 25, 2023, 6:20:03 AM1/25/23
to
Hi Tomas,
On Wed, Jan 25, 2023 at 10:42:50AM +0100, Tomas Pospisek wrote:
> and I guess #1027483 can be closed as it is still open along with #1027430 ?
Thanks for confirming. It does not need to be manually closed, as I
added bug closer in the debian/changelog accordingly. I twill happens
as soon stable release manager accept linux/5.10.162-1 from stable-new
into the stable-proposed-updates .
Regards,
Salvatore
On Wed, Jan 25, 2023 at 10:42:50AM +0100, Tomas Pospisek wrote:
> and I guess #1027483 can be closed as it is still open along with #1027430 ?
added bug closer in the debian/changelog accordingly. I twill happens
as soon stable release manager accept linux/5.10.162-1 from stable-new
into the stable-proposed-updates .
Regards,
Salvatore
0 new messages