I believe the patch you are referring to was related to Asus throttle thermal policy in module asus-wmi. That problem is unrelated to this one. Kernel 6.12.6 (latest stable as of 12/23/24) does not solve this problem. > On Dec 22, 2024, at 5:37 PM, Stuart Balfour <sbalfour@xxxxxxx> wrote: > > I am on kernel 6.12.1, though I downloaded asus-wmi.c from top of tree at git. If you are referring to the patch in Sept. 2024 re OOBE near the bottom of asus_wmi_led_init, yes I rebuilt and tested the driver on the Asus H7606 platform, and it did not enable the backlight. That’s a related problem. Can you provide a URL to the patch submittal you refer to? Specifics really help here. > Thanks. > >> On Dec 22, 2024, at 1:20 PM, Armin Wolf <W_Armin@xxxxxx> wrote: >> >> Am 20.12.24 um 20:28 schrieb Stuart Balfour: >> >>> I have an Asus Proart P16 model H7606WV Ryzen laptop. On Mint 22, kernel 6.12.1 (but later kernels are similarly afflicted), on bootup the hotkey functions of keys F1-F12 are inaccessible. Evtest for event7 (hotkeys) shows no events generated for key press and release F1-F12. Fnlock toggle Fn+ESC is inoperative. Fn+Fxx is similarly inoperative. The problem is that the hotkeys are disabled in the hardware. They work on Windows 11 24H2. The relevant drivers there are contained in the Asus proprietary crapware. If you don’t install the crapware, or uninstall it, you may not have working hotkeys. On Linux, an undefined (asus-wmi.h) ACPI device id is associated with hotkeys and that device is undetectible to the software - querying it returns a presence bit of “0”. It does in fact exist - it took me weeks to find it in the DSDT. Even experts could stump their toes on this one. The putative device id is 0x100013. Asus released this platform mid-Aug. 2024. It’s likely other newer platforms are similarly afflicted. >>> >>> I’m close to a patch in asus-wmi.c, adjacent to the recent patch for toggling the OOBE device during driver initialization. But the device this time cannot be detected. A “secret” device. I think this fix is going to require a change to the ACPI. Poking an unknown device is a recipe for unpredictable effects. First I need to get this reported as a bug. Where would that be? Then how do we fix something that actually requires a custom platform-specific ACPI? >>> >>> Thanks for your attention. It’s been a long haul to get here. >> >> Hi, >> >> the latest 6.12 stable kernel contains an important fix for asus-wmi. Did you test it? >> >> Thanks, >> Armin Wolf >>
