Fedora Account System
Red Hat Associate
Red Hat Customer
1. Please describe the problem: Audio is completely unavailable on Fedora 44 Atomic (Sericea) on a Dell XPS 14 DA14260 after upgrading the kernel. Hardware: Intel Core Ultra Processors (Series 3) HD Audio PCI ID: 8086:e428 Dell subsystem: 1028:0db9 Kernel driver in use: sof-audio-pci-intel-ptl With kernel 7.1.12-200.fc44.x86_64, ALSA registers the sof-soundwire card normally and exposes Jack Out, Speaker, and HDMI playback devices. With kernel 7.2.4-200.fc44.x86_64: cat /proc/asound/cards returns: --- no soundcards --- and: aplay -l returns: aplay: device_list:279: no soundcards found... The most significant difference in the kernel logs is the probing of the four Cirrus Logic CS35L57 SoundWire amplifiers. On the working 7.1.12 kernel, all four devices initialize successfully and report: Cirrus Logic CS35L57 Rev B2 OTP1 fw:4.2.1 (patched=0) The kernel then continues loading the SoundWire topology, applies amplifier calibration, and registers the sof-soundwire sound card. On the broken 7.2.4 kernel, all four CS35L57 amplifier instances fail during probe with messages such as: error -EBUSY: Failed to get spk-id-gpios probe with driver cs35l56 failed with error -16 The kernel also reports: No SoundWire machine driver found for the ACPI-reported configuration: link 0 mfg_id 0x01fa part_id 0x4245 version 0x3 link 2 mfg_id 0x01fa part_id 0x3557 version 0x3 link 3 mfg_id 0x01fa part_id 0x3557 version 0x3 followed by: Use SoundWire default machine driver with function topologies However, the working 7.1.12 kernel also reports this SoundWire machine-driver fallback and successfully initializes the audio hardware, so this warning by itself does not appear to be the regression. The apparent regression is that the CS35L57 amplifiers successfully probe on 7.1.12 but fail with "Failed to get spk-id-gpios" / error -16 on newer kernels, after which no ALSA sound card is registered. 2. What is the Version-Release number of the kernel: Currently broken: 7.2.4-200.fc44.x86_64 Also confirmed broken previously: 7.1.13-200.fc44.x86_64 Last known working: 7.1.12-200.fc44.x86_64 3. Did it work previously in Fedora?: Yes. Audio works normally on: 7.1.12-200.fc44.x86_64 The issue first appeared for me after upgrading to: 7.1.13-200.fc44.x86_64 I no longer have the 7.1.13 deployment available for logs because I updated to 7.2.4 before realizing the regression would persist. The issue remains present on 7.2.4-200.fc44.x86_64. Booting the preserved 7.1.12 deployment still restores audio immediately with no other configuration changes. 4. Can you reproduce this issue? If so, please provide the steps to reproduce the issue below: Yes, consistently. 1. Boot Fedora 44 with kernel 7.2.4-200.fc44.x86_64. 2. Run: cat /proc/asound/cards Result: --- no soundcards --- 3. Run: aplay -l Result: aplay: device_list:279: no soundcards found... 4. Reboot into the preserved 7.1.12-200.fc44.x86_64 deployment. Audio works normally again, and sof-soundwire is registered without making any other configuration changes. 5. Does this problem occur with the latest Rawhide kernel? Not tested. This is an rpm-ostree based Fedora Atomic system, and I did not rebase or otherwise convert the system to Rawhide for this test. 6. Are you running any modules that not shipped with directly Fedora's kernel?: No third-party kernel modules are being used for the audio hardware. The audio device is using Fedora's: sof-audio-pci-intel-ptl with kernel modules including: snd_sof_pci_intel_ptl snd_hda_intel 7. Please attach the kernel logs. Attached Reproducible: Always
Created attachment 2157374 [details] working Journal
Created attachment 2157375 [details] Broken Journal
Created attachment 2157376 [details] Working Dmesg
Created attachment 2157377 [details] Broken Dmesg
I tested Fedora kernel 7.2.5-200.fc44.x86_64 using an rpm-ostree kernel override. The regression is still present. uname -r: 7.2.5-200.fc44.x86_64 cat /proc/asound/cards: --- no soundcards --- aplay -l: aplay: device_list:279: no soundcards found... The kernel log shows the same failure seen on 7.2.4. All four Cirrus Logic CS35L57 SoundWire amplifier instances fail during probe with: error -EBUSY: Failed to get spk-id-gpios probe with driver cs35l56 failed with error -16 The SOF firmware itself loads and boots successfully: Loaded firmware library: ADSPFW, version: 2.14.1.1 Booted firmware version: 2.14.1.1 So the issue remains reproducible on 7.2.5 and appears unchanged from 7.2.4.
Same failure as BZ 2529031 (same model, same four -EBUSY lines); analysis and the fix are there. Your attachments show the order. Broken 7.2.4 dmesg (attachment 2157377 [details]): [13.459885] intel_cvs i2c-INTC10E1:00: Quirks: 0x7a (VID:0x06cb PID:0x0701) [14.796721] cs35l56 sdw:0:2:01fa:3557:01:2: error -EBUSY: Failed to get spk-id-gpios (three more within 2 ms) Working 7.1.12 dmesg (2157376) has no intel_cvs at all. The driver first shipped in Fedora with 7.1.13-200 (config enable plus backport) and is upstream from 7.2, which matches your 7.1.12 good / 7.1.13, 7.2.4, 7.2.5 bad. It claims its wake GPIO on probe; the evidence in BZ 2529031 points at that line as the amps' speaker-ID input, pending the DSDT. Patch that takes the wake IRQ without claiming the line: https://lore.kernel.org/r/20260908105717.496232-1-junjie.cao@intel.com Quick check on 7.2.5: boot with modprobe.blacklist=intel_cvs and confirm the card comes back. An acpidump attachment would let me confirm the pin from the DSDT. Suggest marking this a duplicate of BZ 2529031.
Agreed, all progress seems to be made on BZ-2529031, will mark this as duplicate. Will test 2.4.6 when available. Thank you for your assistance on both threads. *** This bug has been marked as a duplicate of bug 2529031 ***