SND_BCM2835 audio not working on 4.19 kernel version #3181

NguyenTrongThinh · 2019-08-24T08:50:39Z

Describe the bug
Build 64 bit kernel with these configuration
CONFIG_SND_BCM2835=y
CONFIG_VIDEO_BCM2835=y
CONFIG_VIDEO_CODEC_BCM2835=y
but the on board audio (Jack 3.5) was not working, when run command $ aplay -l; it shows no audio card

To reproduce
Build 64 bit kernel with the configuration above

Expected behaviour
Audio on Jack 3.5 can work

Actual behaviour
No audio out and no sound card found when run command aplay -l

System
Raspberry Pi 3 B+
Host Machine: Ubuntu 18.04

6by9 · 2019-08-24T09:09:58Z

Which defconfig did you start from before seting those 3 modules as built in?

bcmrpi3_defconfig already has those drivers defined as being modules. Is there a specific need to make them built in? Did it work for you with them as modules?
Is there anything useful in the kernel log? Please provide the log.

NguyenTrongThinh · 2019-08-24T10:57:25Z

bcmrpi3_defconfig already has those drivers defined as being modules.
=> Yes. default they are "m"

Is there a specific need to make them built in?
=> No, I have changed them to test; because "m" not work and aplay -l shows very strange output. I am sure that I did not connect raspberry PI to any HDMI monitor; And No HDMI cable plugin
$ aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: vc4hdmi [vc4-hdmi], device 0: MAI PCM vc4-hdmi-hifi-0 []
Subdevices: 1/1
Subdevice #0: subdevice #0

Did it work for you with them as modules?
=> No, below are dmesg log and I also change the config to "m" in this log

dmesg.txt

defconfig_rpi3b_plus.txt

config.txt

sakaki- · 2019-08-24T15:57:49Z

You appear to be using vc4-kms-v3d (in /boot/config.txt): does the problem still arise with vc4-fkms-v3d?

asavah · 2019-08-24T21:20:08Z

Out of curiosity I tested this on my own build of 4.19.y arm64:

with no overlay or with vc4-fkms-v3d -> works
with vc4-kms-v3d -> I get same output as op.

NguyenTrongThinh · 2019-08-25T03:42:10Z

@sakaki- @asavah
Thank you for your answer! It is correct.

So now the on board audio card appear but HDMI audio is still there. I am sure that no HDMI cable is plugged in

aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: ALSA [bcm2835 ALSA], device 0: bcm2835 ALSA [bcm2835 ALSA]
Subdevices: 7/7
Subdevice #0: subdevice #0
Subdevice #1: subdevice #1
Subdevice #2: subdevice #2
Subdevice #3: subdevice #3
Subdevice #4: subdevice #4
Subdevice #5: subdevice #5
Subdevice #6: subdevice #6
card 0: ALSA [bcm2835 ALSA], device 1: bcm2835 IEC958/HDMI [bcm2835 IEC958/HDMI]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 0: ALSA [bcm2835 ALSA], device 2: bcm2835 IEC958/HDMI1 [bcm2835 IEC958/HDMI1]
Subdevices: 1/1
Subdevice #0: subdevice #0

floion · 2019-08-29T13:03:48Z

I confirm the same bug on 4.19.66. On 4.19.58 the bug is not present.
Using vc4-fkms-v3d instead of vc4-kms-v3d as @sakaki- and @asavah pointed out above, the problem is fixed on 4.19.66

6by9 · 2019-08-29T13:29:42Z

vc4-hdmi is the upstream audio driver provided as part of the full DRM/KMS driver (vc4-kms-v3d).
The HDMI peripheral only accepts IEC958 formatted data, hence your output

$ aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: vc4hdmi [vc4-hdmi], device 0: MAI PCM vc4-hdmi-hifi-0 []
Subdevices: 1/1
Subdevice #0: subdevice #0

In theory you can get ALSA to do the relevant repacking of the data, but upstream never seemed to fully resolve the issues. See #2988

Using snd_bcm2835 (dtparam=audio=on) will expose one audio device that can be switched between multiple outputs (audio, HDMI0, and HDMI1 (on a Pi4 only)), and also HDMI passthrough subdevices. The HDMI passthrough devices are not runtime checked - they will always be present.

lategoodbye · 2019-08-29T17:29:17Z

There is no relation between vc4-kms-v3d and snd_bcm2835. But maybe there is some kind of regression in the VCHIQ driver which prevent probing of snd_bcm2835 in case vc4-fkms-v3d is not probed.

asavah · 2019-08-29T17:57:52Z

@lategoodbye With no *kms overlay at all snd_bcm2835 is probed and works correctly.

Note that dtoverlay is commented out in config.txt:

grep kms /boot/config.txt && aplay -l
#dtoverlay=vc4-fkms-v3d
**** List of PLAYBACK Hardware Devices ****
card 0: ALSA [bcm2835 ALSA], device 0: bcm2835 ALSA [bcm2835 ALSA]
  Subdevices: 7/7
  Subdevice #0: subdevice #0
  Subdevice #1: subdevice #1
  Subdevice #2: subdevice #2
  Subdevice #3: subdevice #3
  Subdevice #4: subdevice #4
  Subdevice #5: subdevice #5
  Subdevice #6: subdevice #6
card 0: ALSA [bcm2835 ALSA], device 1: bcm2835 IEC958/HDMI [bcm2835 IEC958/HDMI]
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 0: ALSA [bcm2835 ALSA], device 2: bcm2835 IEC958/HDMI1 [bcm2835 IEC958/HDMI1]
  Subdevices: 1/1
  Subdevice #0: subdevice #0

kernel 4.19.68 arm64

floion · 2019-09-05T06:04:56Z

@6by9 so is this considered a bug after all?

6by9 · 2019-09-05T07:08:07Z

64 bit kernels are in the early stages still. I haven't had a chance to look as most of my time is on Pi4 at present.

rpi-update does now include a 64bit kernel, admittedly aimed more at Pi4 than Pi3, so retesting with a standard kernel would be useful. See https://www.raspberrypi.org/forums/viewtopic.php?f=29&t=250730 for more details.

lategoodbye · 2019-09-05T08:44:11Z

Looking at the dmesg from @NguyenTrongThinh shows there is no attempt to load snd_bcm2835.
Can someone how is affected verify that the status of snd_bcm2835 devicetree node is really set to "okay"?

6by9 · 2019-09-05T09:55:15Z

I've rpi-updated by Pi3, added arm_64bit=1, dtparam=audio=on, and dtoverlay=vc4-kms-v3d to config.txt.

pi@raspberrypi:~ $ aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: vc4hdmi [vc4-hdmi], device 0: MAI PCM vc4-hdmi-hifi-0 []
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 1: ALSA [bcm2835 ALSA], device 0: bcm2835 ALSA [bcm2835 ALSA]
  Subdevices: 7/7
  Subdevice #0: subdevice #0
  Subdevice #1: subdevice #1
  Subdevice #2: subdevice #2
  Subdevice #3: subdevice #3
  Subdevice #4: subdevice #4
  Subdevice #5: subdevice #5
  Subdevice #6: subdevice #6
card 1: ALSA [bcm2835 ALSA], device 1: bcm2835 IEC958/HDMI [bcm2835 IEC958/HDMI]
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 1: ALSA [bcm2835 ALSA], device 2: bcm2835 IEC958/HDMI1 [bcm2835 IEC958/HDMI1]
  Subdevices: 1/1
  Subdevice #0: subdevice #0

omxplayer plays audio happily, so the VPU can produce sound (bypasses alsa).
VLC plays audio happily if directed at [bcm2835 ALSA] default device (it can't support the IEC958 format required by vc4-hdmi).
aplay -v -D plug:iec958 foo.wav plays happily through vc4-hdmi, but having done that snd_bcm2835 is distorted, probably because the two drivers are now fighting for the audio part of the HDMI peripheral.

Disable vc4-hdmi audio (dtoverlay=vc4-kms-v3d,audio=off) with dtparam=audio=on, and we have the full KMS driver with snd_bcm2835.

I fail to see any form of bug there. Back to those reporting it to give more details of what they are doing differently.

floion · 2019-09-05T12:46:01Z

I have the following:

vcgencmd version
Nov  4 2018 16:31:07 
Copyright (c) 2012 Broadcom
version ed5baf9520a3c4ca82ba38594b898f0c0446da66 (clean) (release)

My config.txt is the following (I'm on 32 bits):

dtoverlay=vc4-kms-v3d,audio=off
dtparam=i2c_arm=on
dtparam=spi=on
dtparam=audio=on
disable_splash=1
avoid_warnings=1

and aplay -l is:

**** List of PLAYBACK Hardware Devices ****
card 0: vc4hdmi [vc4-hdmi], device 0: MAI PCM vc4-hdmi-hifi-0 []
  Subdevices: 1/1
  Subdevice #0: subdevice #0

6by9 · 2019-09-05T12:59:55Z

vcgencmd version
Nov 4 2018 16:31:07

Why such an ancient firmware?
What kernel version (uname -a)? You previously mentioned 4.19.66 which was 9th Aug 2019. 4.19.69 is the latest on this branch.

And if that is what aplay -l is reporting, then the audio=off part of dtoverlay=vc4-kms-v3d is not working. Where have your device tree files come from? They obviously don't include e8e84b7

floion · 2019-09-05T13:27:54Z

Updated the kernel again to 4.19.66 (I had temporarily downgraded) and indeed, I do have that commit included. So now if I'm disabling hdmi audio as you suggested, I do have the devices listed by aplay and audio playback works as expected.

Thanks for all the input, much appreciated.

floion · 2019-09-06T09:14:00Z

Sorry to revisit this, seems there is something strange going on. Can't make it work today anymore. Wondering if there was any obscure config.txt setting I had that made it work.

vcgencmd version
Aug 28 2019 15:08:27 
Copyright (c) 2012 Broadcom
version 31098ceed842ca10c8ffd02ffe3640bf6c0190b4 (clean) (release) (start)

uname -a
Linux 68d00bf 4.19.66 #1 SMP Thu Sep 5 21:09:51 UTC 2019 armv7l GNU/Linux

config.txt:

avoid_warnings=1
disable_splash=1
dtoverlay=vc4-kms-v3d,audio=off
dtparam=i2c_arm=on
dtparam=spi=on
dtparam=audio=on
gpu_mem=396

I do have e8e84b7 applied.

aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: vc4hdmi [vc4-hdmi], device 0: MAI PCM vc4-hdmi-hifi-0 []
  Subdevices: 1/1
  Subdevice #0: subdevice #0

Wondering if somebody else can reproduce this

6by9 · 2019-09-06T09:35:22Z

If you have vc4-hdmi ALSA device then somewhere you haven't disabled audio. Check for the file /proc/device-tree/soc/hdmi@7e902000/dmas.

It should be 0 bytes in size if audio=off has applied correctly.

If still enabled then I'd expect it to be 8 bytes

pi@raspberrypi:~ $ xxd /proc/device-tree/soc/hdmi@7e902000/dmas
00000000: 0000 000a 0000 0011

For snd_bcm2835, check /proc/device-tree/soc/audio/status contains the string "okay" (as lategoodbye commented yesterday).

warpme · 2019-11-12T13:54:46Z

Hi,
I'm using current rpi kernel with mainline U-boot and no matter what I can't get sound on RPI3 as I'm always ending with vc4-hdmi upstream audio driver. I already tried this comment but I'm constantly getting like this comment + there is no audio in /proc/device-tree/soc
It looks like entries in config.txt are ignored...
May somebody hint me how to move forward with this?

pelwell · 2019-11-12T13:57:46Z

Unless U-boot is configured to pass through the firmware-supplied DTB, the DTB will be loaded from scratch by U-boot. This will lose any overlays and parameters set in config.txt.

warpme · 2019-11-12T18:57:30Z

hmm,
here is what i done:

build my OS for armv7 arch
built mainline u-boot with rpi_3_32b_defconfig
copy u-boot.bin to /boot
add kernel=u-boot.bin to config.txt
prepare boot.scr with this:

fatload mmc 0:1 ${kernel_addr_r} kernel7.img
setenv bootargs earlyprintk dwc_otg.lpm_enable=0 console=serial0,115200 root=/dev/mmcblk0p2 rootfstype=ext4 rootwait
bootz ${kernel_addr_r} - ${fdt_addr}

copy boot.scr to /boot
use kernel7.img from here

powering-on rpi3 gives me u-boot initial u-boot log entries on screen and then all hangs on "Loading kernel...."

probably is still missing something :-(

so Q is: how to achieve

"U-boot configured to pass through the firmware-supplied DTB"
?

pelwell · 2019-11-12T21:10:30Z

The firmware treats U-boot as a kernel and passes it a pointer to the DTB it has constructed. If U-boot and boot.scr are correctly configured I believe the firmware's DTB can be passed on to the Linux kernel, but you are more likely to find help for your problem on a U-boot forum.

JamesH65 · 2019-12-20T15:40:43Z

This issue will be closed within 30 days unless further interactions are posted. If you wish this issue to remain open, please add a comment. A closed issue may be reopened if requested.

NguyenTrongThinh changed the title ~~SND_BCM2835 audio not working on latest kernel version~~ SND_BCM2835 audio not working on 4.19 kernel version Aug 24, 2019

sakaki- mentioned this issue Aug 24, 2019

Audio does not work on RPI 3 B+ (Jack 3.5) sakaki-/gentoo-on-rpi-64bit#101

Closed

tiagoboldt mentioned this issue Aug 31, 2019

Pi4 No audio from HDMI #3197

Closed

pdcastro mentioned this issue Sep 3, 2019

RPi 3 audio and hdmi rotation possibly affected by 2.41.0 RPi 4 updates balena-os/balena-raspberrypi#379

Closed

NguyenTrongThinh closed this as completed Dec 23, 2019

MichaIng mentioned this issue Jan 24, 2020

Amiberry fast boot doesn't wait for sound / issue with RPi 4 sound MichaIng/DietPi#3356

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

SND_BCM2835 audio not working on 4.19 kernel version #3181

SND_BCM2835 audio not working on 4.19 kernel version #3181

NguyenTrongThinh commented Aug 24, 2019

6by9 commented Aug 24, 2019

NguyenTrongThinh commented Aug 24, 2019 •

edited

Loading

sakaki- commented Aug 24, 2019

asavah commented Aug 24, 2019 •

edited

Loading

NguyenTrongThinh commented Aug 25, 2019

floion commented Aug 29, 2019

6by9 commented Aug 29, 2019

lategoodbye commented Aug 29, 2019

asavah commented Aug 29, 2019 •

edited

Loading

floion commented Sep 5, 2019

6by9 commented Sep 5, 2019

lategoodbye commented Sep 5, 2019

6by9 commented Sep 5, 2019

floion commented Sep 5, 2019

6by9 commented Sep 5, 2019 •

edited

Loading

floion commented Sep 5, 2019

floion commented Sep 6, 2019

6by9 commented Sep 6, 2019

warpme commented Nov 12, 2019 •

edited

Loading

pelwell commented Nov 12, 2019 •

edited

Loading

warpme commented Nov 12, 2019 •

edited

Loading

pelwell commented Nov 12, 2019

JamesH65 commented Dec 20, 2019

SND_BCM2835 audio not working on 4.19 kernel version #3181

SND_BCM2835 audio not working on 4.19 kernel version #3181

Comments

NguyenTrongThinh commented Aug 24, 2019

6by9 commented Aug 24, 2019

NguyenTrongThinh commented Aug 24, 2019 • edited Loading

sakaki- commented Aug 24, 2019

asavah commented Aug 24, 2019 • edited Loading

NguyenTrongThinh commented Aug 25, 2019

floion commented Aug 29, 2019

6by9 commented Aug 29, 2019

lategoodbye commented Aug 29, 2019

asavah commented Aug 29, 2019 • edited Loading

floion commented Sep 5, 2019

6by9 commented Sep 5, 2019

lategoodbye commented Sep 5, 2019

6by9 commented Sep 5, 2019

floion commented Sep 5, 2019

6by9 commented Sep 5, 2019 • edited Loading

floion commented Sep 5, 2019

floion commented Sep 6, 2019

6by9 commented Sep 6, 2019

warpme commented Nov 12, 2019 • edited Loading

pelwell commented Nov 12, 2019 • edited Loading

warpme commented Nov 12, 2019 • edited Loading

pelwell commented Nov 12, 2019

JamesH65 commented Dec 20, 2019

NguyenTrongThinh commented Aug 24, 2019 •

edited

Loading

asavah commented Aug 24, 2019 •

edited

Loading

asavah commented Aug 29, 2019 •

edited

Loading

6by9 commented Sep 5, 2019 •

edited

Loading

warpme commented Nov 12, 2019 •

edited

Loading

pelwell commented Nov 12, 2019 •

edited

Loading

warpme commented Nov 12, 2019 •

edited

Loading