Audio driver fails to load on HP Omen 14, Intel Core 9 Ultra Meteor Lake #4880
Comments
|
The sof-mtl-rt711.tplg was only reported as it did not detect the actual hardware. It would be useful if you could find the actual chipset used by the laptop. Alternatively, try signing and booting the latest 6.9 kernel. |
|
This laptop is genuinely using Realtek parts: [ 15.856504] soundwire_bus:sdw_extract_slave_id: soundwire sdw-master-0-0: SDW Slave Addr: 30025d071101 So completely unrelated issues to the other ticket. |
|
Yes, it's a different issue indeed. Thanks @CNG for the detailed reports, much appreciated. That's a new one, not sure what's going on, but unfortunately every new generation of HP devices seems to have issues with the ACPI definitions. We do have existing quirks for HP 'Omen' devices, but maybe they are not applied correctly. see e.g. #4088 for this sort of problems can you provide the results of this command that will tell us what devices are listed in ACPI and then we can check the quirks. Thanks! |
|
Sure thing! |
|
ok so that's rt711-sdca on link0 and rt1316 on link3. Of course we don't have this configuration with a single amp, we only have the configuration with two amps... |
This platform has an RT711-sdca on link0 and RT1316 on link3 Closes: thesofproject#4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
|
@CNG can you apply this patch on your kernel and retest? That should detect the right configuration, but we still need to provide an updated topology file and also figure out why the link0 is not started. |
|
Also unzip this file generated from thesofproject/sof#8991 Completely untested, just to try to see what happens. |
|
Had to read a bit of documentation (not having patched kernel before), and waiting for the package to build now. In case you get this before that's done, just to clarify, you'd like me to boot into the patched kernel and also add that topology file as part of the same test, or should I test each thing individually and together for 3 tests? For each test, would I be providing a new "alsa-info" or running other commands too? (I believe that alsa-info already contained the dmesg output, for example.) For my own reference, here's how I'm attempting to patch and build the kernel. Hopefully I've read the docs correctly: |
|
I am not familiar at all with Arch Linux and always sucked at pacman. Maybe @ujfalusi can help? We have instructions here to build a kernel: https://thesofproject.github.io/latest/getting_started/setup_linux/prepare_build_environment.html |
|
Thanks, it actually just finished, so I'll attempt to boot into it now. |
|
All right, let me know if these look right. No sound magically appeared of course, but hopefully something useful. Also note I still have that
Also I see my kernel got bumped up from |
|
ok, that's good progress @CNG. The remaining issue is the topology, we have an error That's not really surprising, I don't really understand all those DMIC options so not sure what the right settings should be. Please try with this file where I used only the defaults |
|
All right, that produced:
|
|
ok, now the HDMI index is wrong... try this |
|
All right, well this is exciting, I've got new devices showing up now. First here is the output: I don't think I can get sound on either the speakers or headphones. I tried playing music through both "Meteor Lake-P HD Audio Controller Speaker" and "Meteor Lake-P HD Audio Controller Headphones" using both profiles "Play HiFi quality Music" and "Pro Audio". I think I may hear a little periodic static clicks through the headphones, not sure if it is anything. |
|
at this point the kernel is behaving as normal with a card created. the only things that could be problematic is userspace (UCM specifically). Let us know how this goes. |
|
Strangely, when I opened pavucontrol and toggled the configuration for this card to "Pro Audio" and then back to "Play HiFi quality Music", it took a while for the Output Devices tab to reflect the change. Then I played with the volume slider on the "Speaker" item, and got some weird sound out of the speakers. It sounds like it's the normal pop sound effect of moving the volume slider but it's distorted and played at a long delay, my speakers were making noise for more than 30 seconds after I moved the slider. Then trying to play music with Spotify, it actually thinks it can play music... but I don't hear anything from the speaker except for the popping sound from moving the volume slider? |
|
That UI is also reporting headphones are "plugged in" regardless of whether that's true, fwiw. |
|
You probably want to check with a terminal if you can get a clear sine wave with And then any sort of PCM .wav file with aplay -Dhw:0,2 userspace issues are not strictly speaking SOF driver issues, so they might need to be addressed somewhere else. |
|
for the jack detection, i'd bet they used JD1 instead of the default JD2... That would require another kernel patch. |
|
All right, well for the
|
|
try |
|
|
|
that's because your UI is active or something in userspace is keeping the sound server alive. kill all your apps and UIs, let the sound server suspend and then try with a terminal. it's definitively hw:1,2 based on your alsa-info |
|
Ah you're right. Even with everything closed I was getting the "resource busy" message, but now after rebooting it works better. I am getting a sine sound out of both speakers when it says "Front Left", and getting no sound at all when it says "Front Right". I'm not yet getting output testing |
|
You probably have an invalid ALSA mixer configuration. Usually those are provided via UCM, the file alsa-ucm-conf/ucm2/sof-soundwire/rt1316.conf seems to select the correct mixers for the single amp case. check if your distro has the latest stuff from UCM upstream: For the aplay test, you have to use a 48kHz 2ch file. We don't support mono or other rates. |
This device has RT711 on link0, RT1316 on link3 and 2 PCH-attached DMICs. Add 3 configurations with no DMIC, 4 DMIC and 2 DMIC. For now no support for -pdm1. Link: thesofproject/linux#4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
|
For the jack detection issue, please try this additional patch on top of what you already have: https://github.com/thesofproject/linux/commit/e1ff44dfed3e4c450b96617c9ba06dd1d76da703.patch If this doesn't work, try with JD2_100K instead of JD2. We have no way of discovering how the jack detection is handled so a quirk is needed if the defaults don't work. |
|
OK regarding the ALSA mixer config, I did note there are some changes between my local file and the Github version you linked. I did not yet trace how that file is getting to my system etc. but for now manually updated it: After rebooting, I reran the I'll work on testing that second kernel patch next. Small note, the file is provided by my distro's |
|
The UCM update is indeed very recent, see the log below. It's not really surprising it's not in any distribution just yet. FWIW most distros don't have the latest stuff needed for Meteor Lake, it's still a rather new platform where the variations chosen by OEMs are not yet known. |
|
All right, some progress. Using your last patch, the jack detection seems to work now. Opening I tried to play sound through the headphones using I assume something crashed, but not sure what to look for. As a stab in the dark I tried: Anyway, I'm not sure at this point what pieces you can help with, where I should seek help elsewhere, what is likely already being addressed by others if I were more patient, etc. I just want to be of service in case something wouldn't otherwise be addressed, if that makes sense. Thank you! |
|
ok, thanks for testing the jack detection. That's a good sign. for the rest, it's where remote debug becomes difficult. there are just too many moving parts with userspace components that are not all well aligned. If it was my own device, I would try to make the sound work on all endpoints with alsa-ucm in a command line. this should be something like alsaucm -c sof-soundwire set _verb HiFi set _enadev "Jack Out" and then check what happens if you modify the mixer settings with something like (possible syntax or typos, but you should get the idea) amixer -Dhw:1 cset name='rt711 Headphone Switch' on alsaucm -c sof-soundwire set _verb HiFi set _enadev Speaker amixer -Dhw:1 cset name='rt1316-1 RX Channel Select' "L/R" |
This platform has an RT711-sdca on link0 and RT1316 on link3 Closes: #4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
This platform has an RT711-sdca on link0 and RT1316 on link3 Closes: #4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
This device has RT711 on link0, RT1316 on link3 and 2 PCH-attached DMICs. Add 3 configurations with no DMIC, 4 DMIC and 2 DMIC. For now no support for -pdm1. Link: thesofproject/linux#4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
This platform has an RT711-sdca on link0 and RT1316 on link3 Closes: #4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
This platform has an RT711-sdca on link0 and RT1316 on link3 Closes: #4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
This platform has an RT711-sdca on link0 and RT1316 on link3 Closes: thesofproject#4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
This platform has an RT711-sdca on link0 and RT1316 on link3 Closes: thesofproject#4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20240411220347.131267-3-pierre-louis.bossart@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
This device has RT711 on link0, RT1316 on link3 and 2 PCH-attached DMICs. Add 3 configurations with no DMIC, 4 DMIC and 2 DMIC. For now no support for -pdm1. Link: thesofproject/linux#4880 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
|
Should this fix be rolled out fully now? I'm on the same computer using arch linux but do not have audio and get the same dmesg errors as posted in the original post, using kernel 6.10 (was using 6.9 before) and have the latest available sof-firmware 2024.06-1 and have the latest alsa-ucm-conf which is from june 12 (1.2.12) |
|
the patch is in 6.10, see |
|
@ImDevinC do let me know if you get your sound working. :-) I was not able to have success running with Mr. Bossart's March 28 comment on next steps I should try. |
|
@ImDevinC please confirm that you are indeed using 6.10 and if you see the issue with the wrong topology being requested. |
|
@CNG I don't remember where I left my keys, I can't possibly remember what I said in March haha. From the comments it seems the kernel didn't show any bad messages, jack detection worked. Not sure if UCM ever worked? |
|
Ahh my bad, I'm on 6.9.10 (misread the update), will look at 6.10 next week and confirm again. Apologies |
|
Just confirmed I'm on 6.10 and while the audio device shows up now, I do not get any audio. Here's what I'm seeing in journalctl for "sof-". My dmesg buffer is too large to go back to the beginning because there is a bunch of spam messages about the audio card (I'll put those below) Repeated messages in dmesg Using the command above, |
|
This seems to be a different issue @ImDevinC, the topology is found and the configuration is ok. The problem now is that the firmware has some sort of DSP panic. It seems you have the latest version that was released last week, do you mind trying with the SOF 2.9 firmware just to rule out a firmware issue? a DSP panic is not expected... |
|
forked issue to #5116, let's continue there. |
Hello,
On my HP OMEN Transcend Gaming Laptop 14t-fb000 with an Intel Core 9 Ultra CPU I have a similar driver failure issue as #4879 but my topology file that's missing
sof-mtl-rt711-2ch.tplgis slightly different than @moritz89'ssof-mtl-rt711.tplg. I initially wrote this up on the Arch Linux forums at No builtin speaker audio, HP Omen 14, Intel Core 9 Ultra Meteor Lake but haven't gotten any feedback there. Hopefully this is a more appropriate place to seek guidance.Environment
sudo dmesg | grep -E "sof-|snd"I do have Arch's latest
sof-firmwareandsof-binpackages installed. I also tried to locate that referencedsof-mtl-rt711-2ch.tplgtopology file, but did not find it in the sof-bin repo. Indeed the only result on Google for that string is my forum post on this issue.Per recommendation in #4879 I've saved sof-dyndbg.conf.txt to
/etc/modprobe.d/sof-dyndbg.confand generated alsa-info. I see the dmesg is at the bottom of that, but here is it filtered with the same pattern I had used above for completeness:sudo dmesg | grep -E "sof-|snd"Is there anything else I should provide?
Thank you!
Charlie
The text was updated successfully, but these errors were encountered: