09:24 <lpieralisi> HdkR: PSCI_STAT_RESIDENCY does not track time spent in "plain" WFI since it is not a PSCI power state per-se. For WFE PMU counters should do.

13:54 <krzk> Anyone knows nice example (template) of driver which will be a simple bus provider for clients? I am looking to give some hints to contributor for something really SoC-specific, so without defining bus_type, but rather something simpler.

13:54 <krzk> Or maybe this is discouraged and even simple vendor buses should be a "bus_type"?

13:56 <krzk> Strictly speaking this is about Samsung Speedy which is some type of different and incompatible implementation of Dallas 1-wire. https://lore.kernel.org/all/20241212-speedy-v1-0-544ad7bcfb6a@gmail.com/

14:18 <geertu> krzk: Copy and corrupt^Wmodify drivers/w1?

14:19 <geertu> Any bitbang GPIO implementation?

14:23 <javierm> krzk: I would grep for of_platform_populate() in drivers/

14:53 <ajb-linaro> the command echo 'file arch/arm64/kernel/alignment.c +p' /proc/dynamic_debug/control should enable all pr_debug() entries for a given file right?

15:07 <HdkR> lpieralisi: Alright cool, so I'll need to look in to PMU counters

15:31 <ajb-linaro> hmm all my amdgpu problems go away if I blacklist the module and manually load it once the system has booted

15:31 <ajb-linaro> I guess it wasn't the alignment hacks after all