Update with additional testing on kernel 7.0.0-34-generic (BIOS F.23). Correction to the original report: COEF registers DO respond to hda-verb (I could read and write COEF index 0x10 on node 0x20), so that statement in the description is wrong.
Findings: 1. With snd_hda_codec.dyndbg=+p the kernel reports: "ALC295: picked fixup for PCI SSID 103c:0000" i.e. a generic HP fallback (empty name); there is no entry for 103c:86e3. 2. Codec GPIOs (/proc/asound/card1/codec#0): GPIO io=3, and IO[0..2] all enable=0 dir=0 data=0. Nothing enables them. 3. ACPI: I disassembled DSDT and all SSDTs (acpidump/iasl). A case-sensitive search for LED, MLED, MUTL and "mute" finds no matching symbols. WMID.GWEV calls are HP WMI events only. 4. Manual test of what ALC295_FIXUP_HP_MUTE_LED_COEFBIT11 does: setting bit 11 of COEF 0x10 on node 0x20 (0x20 -> 0x820, then restored) does NOT light the LED on this machine. So that quirk alone is not enough. 5. Previously tried without success: alc295-hp-x360, hp-mute-led-mic3, HDA Jack Retask, Intel SOF (dsp_driver=3), raw hda-verb on GPIO/COEF. hp_wmi is loaded. Not concluded: runtime tests of the other hp-mute-led-* models via modelname/reconfig were inconclusive, so I don't consider them ruled out. I don't know whether another COEF bit or the Embedded Controller drives the LED. The LED works on Windows, so a driver-level method exists. If anyone knows which register/interface HP uses, I can test patches and report back. Attached: dmesg-hda.txt, alsa-info, DSDT/SSDT tables (MSDM removed). -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2155632 Title: Mute LED not working on HP Pavilion 15-cs3xxx (ALC295, subsystem 0x103c86e3) To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2155632/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
