amdgpu_devic...gating_state(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… amdgpu_device_ip_get_clockgating_state β€” the kernel API manual, but make it make sense

okay so bestie, welcome to the glow-up edition of amdgpu_devic...gating_state(9) from the Kernel API Manual. yes the title is truncated with dots in the header, that's just how the manpage stylists do it, don't come for them πŸŽ€


✨ NAME ✨

amdgpu_device_ip_get_clockgating_state β€” get the CG state 🫢

like, literally just fetch the clockgating state. that's the whole vibe.


🌸 SYNOPSIS 🌸

void amdgpu_device_ip_get_clockgating_state (struct amdgpu_device *adev
, u64 *flags );

it returns void, so don't go looking for a return value, bestie, there isn't one πŸ’€


πŸ¦‹ ARGUMENTS πŸ¦‹

  • adev β€” an amdgpu_device pointer. yeah. the device. the main character of this whole situation πŸ‘‘
  • flags β€” the clockgating feature flags. you hand it over and it gets updated, iykyk.

πŸ”₯ DESCRIPTION πŸ”₯

so what actually happens? it walks the list of IPs on the device and updates the clockgating flags for each IP. that's it, that's the plot ✨

more specifically, it updates flags with the feature flags for each hardware IP where clockgating is enabled. so the ones that aren't clockgated? not invited to the party πŸ™ˆ the ones that are? they're getting their flags written in, periodt πŸ’…

ngl this one's a getter with side effects on your flags pointer, so be chill about that and make sure you actually want it written to. no cap.

🌈 SEE ALSO 🌈

starting with the kernel file, iconic behaviour:

the whole friend group fr fr πŸ§šπŸ’œ


πŸ’– that's the page!! sourced from ./drivers/gpu/drm/amd/amdgpu, dated October 2026, and yes it's filed under amdgpu_devic...gating_state(9) β€” go read it again if you forgot anything, slay πŸŽ€β­

πŸ“¦ where this page comes from